EVALUATING PUBLIC RECORDS OF SUPPLY TRANSACTIONS
A platform facilitates buyers, sellers, and third parties in obtaining information related to each other's transaction histories, such as a supplier's shipment history, the types of materials typically shipped, a supplier's customers, a supplier's expertise, what materials and how much a buyer purchases, buyer and shipper reliability, similarity between buyers, similarity between suppliers, and the like. The platform aggregates data from a variety of sources, including, without limitation, customs data associated with actual import/export transactions and facilitates the generation of reports as to the quality of buyers and suppliers, the reports relating to a variety of parameters that are associated with buyer and supplier quality.
Latest Panjiva, Inc. Patents:
This application is a continuation of U.S. application Ser. No. 13/906,007 filed May 30, 2013, which is a continuation of U.S. application Ser. No. 12/271,593 filed Nov. 14, 2008 and patented as U.S. Pat. No. 8,473,354 on Jun. 25, 2013, which claims the benefit of U.S. provisional application Ser. No. 60/987,989 filed Nov. 14, 2007, each of which is hereby incorporated by reference in its entirety.
BACKGROUND1. Field
The present invention is related to electronic commerce, and more particularly to rating systems.
2. Description of the Related Art
Buyers who are interested in working with suppliers, particularly overseas suppliers, may have many suppliers from which they can choose. For instance, in the apparel industry there are an estimated 40,000 apparel factories in China alone, with some 80,000 worldwide. In order to select a supplier, a buyer traditionally has had to rely on direct experience with the supplier or work through a middleman that facilitates contracting with suppliers. However, working with a middle-man may incur commissions for their services, and working directly with the supplier may present the buyer with a large degree of uncertainty, such as relating to the quality and reliability of the supplier, who the supplier typically works with, what type of products the supplier typically supplies, materials used, customers served, and the like. Some information about suppliers can be obtained from other sources, such as trade fairs, online directories, referrals, and the like. These disparate sources of information are, however, difficult to sort through, and at present there is a distinct lack of reliable and objective information that buyers can use to assess suppliers around the world. As a result, buyers must proceed largely on their own, and at considerable risk and expense.
A need exists for ways for buyers to more easily select suppliers.
SUMMARYMethods and systems are disclosed herein for a platform by which buyers, sellers, and third parties can obtain information related to each other's transaction histories, such as a supplier's shipment history, the types of materials typically shipped, a supplier's customers, a supplier's expertise, what materials and how much a buyer purchases, buyer and shipper reliability, similarity between buyers, similarity between suppliers, and the like. The platform may aggregate data from a variety of sources, including, without limitation, customs data associated with actual import/export transactions, and facilitates the generation of reports as to the quality of buyers and suppliers, the reports relating to a variety of parameters that are associated with quality buyers and suppliers, and the like.
In an aspect of the invention, methods and systems may include: using a computer implemented facility to collect and store a plurality of records of transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; associating the transactions with entities; and rating an entity based on analysis of the aggregated transactions. In the aspect a rating is tailored based on criteria defined by an end user.
In the aspect a rating is for one or more of: suppliers using aggregated transactional customs data, a supplier based on customs data related to transactions by the supplier with a third party, a buyer using aggregated transactional customs data, a buyer based on customs data related to transactions of the buyer with a third party, a supplier based on loyalty as indicated by analysis of customs transactions, a supplier based on amount of experience as indicated by customs transactions, a supplier based on evaluating the number of shipments, a supplier based on duration of experience as indicated by shipments, a supplier based on size of transactions as indicated by past shipments, a supplier based on extent of international experience as indicated by past shipments, a supplier based on extent of country-relevant experience as indicated by past shipments, a buyer based on loyalty as indicated by analysis of customs transactions, a buyer based on amount of experience as indicated by customs transactions, a buyer based on evaluating the number of shipments, a buyer based on duration of experience as indicated by shipments, a buyer based on size of transactions as indicated by past shipments, a buyer based on extent of international experience as indicated by past shipments, a buyer based on extent of country-relevant experience as indicated by past shipments, a supplier based on customer loyalty and supplier experience as indicated by past shipments reflected in customs records.
In the aspect, the rating is further based on at least two factors selected from the group consisting of: a country context of a party, a business legitimacy of a party, whether a party is registered with government authorities, an assessment of a trading environment in a country, macroeconomic information, public recognition of a party, industry awards, industry certifications, amount of experience, number of shipments, duration of experience, size of transactions, extent of domestic experience, extent of international experience, caliber of customers, customer loyalty, degree of specialization, specialization in product categories, specialization in manufacturing techniques, specialization in materials, specialization in gender, feedback from customers, feedback from buyers, feedback on product quality, feedback on customer service, feedback on timeliness of delivery, feedback on language skills, feedback on sample making ability, respect for intellectual property, quality management, social responsibility, environmental responsibility, standards of compliance, certifications, and certifications with respect to specific vendor standards.
In the aspect, the rating is based on one of: a country context of a party, a business legitimacy of a party, whether a party is registered with government authorities, an assessment of a trading environment in a country, macroeconomic information, public recognition of a party, industry awards, industry certifications, amount of experience, number of shipments, duration of experience, size of transactions, extent of domestic experience, extent of international experience, caliber of customers, customer loyalty, degree of specialization, specialization in product categories, specialization in manufacturing techniques, specialization in materials, specialization in gender, feedback from customers, feedback from buyers, feedback on product quality, feedback on customer service, feedback on timeliness of delivery, feedback on language skills, feedback on sample making ability, respect for intellectual property, quality management, social responsibility, environmental responsibility, standards of compliance, certifications, and certifications with respect to specific vendor standards
In the aspect, weights are given in the rating process. The weights are based on timeliness of data. The weights are given based on size of transaction. The weights for transactions are given based on the quality of the transacting parties; the quality of a transacting party is based on a prior rating for that party. The weights are based on relevance of data.
In the aspect the rating is for a plurality of factories of an entity.
In the aspect the rating includes providing a human-aided assessment of supplier skills as a factor in a rating. Alternatively the rating includes using an indicator of an entity's financial health as a factor in a rating.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of records of customs transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; associating the transactions with entities; and providing an entity score for an entity based on analysis of the aggregated transactions. In the aspect the entity score is based at least in part on transactional data about shipments by the entity. In the aspect the entity score includes factors selected from the group consisting of country context, business legitimacy information, public recognition, amount of experience, caliber of customers of the supplier, customer loyalty for the supplier, degree of specialization of the supplier, and feedback from previous customers. In the aspect the entity score for the suppliers is based on aggregated transactional customs data. In the aspect the entity score is based on a criteria defined by an end user. In the aspect the entity score for a supplier is based on customs data related to transactions by the supplier with a third party. In the aspect the entity score is based on at least two factors selected from the group consisting of: a country context of a party, a business legitimacy of a party, whether a party is registered with government authorities, an assessment of a trading environment in a country, macroeconomic information, public recognition of a party, industry awards, industry certifications, amount of experience, number of shipments, duration of experience, size of transactions, extent of domestic experience, extent of international experience, caliber of customers, customer loyalty, degree of specialization, specialization in product categories, specialization in manufacturing techniques, specialization in materials, specialization in gender, feedback from customers, feedback from buyers, feedback on product quality, feedback on customer service, feedback on timeliness of delivery, feedback on language skills, feedback on sample making ability, respect for intellectual property, quality management, social responsibility, environmental responsibility, standards of compliance, certifications, and certifications with respect to specific vendor standards. In the aspect the entity score is based upon one of: a country context of a party, a business legitimacy of a party, whether a party is registered with government authorities, an assessment of a trading environment in a country, macroeconomic information, public recognition of a party, industry awards, industry certifications, amount of experience, number of shipments, duration of experience, size of transactions, extent of domestic experience, extent of international experience, caliber of customers, customer loyalty, degree of specialization, specialization in product categories, specialization in manufacturing techniques, specialization in materials, specialization in gender, feedback from customers, feedback from buyers, feedback on product quality, feedback on customer service, feedback on timeliness of delivery, feedback on language skills, feedback on sample making ability, respect for intellectual property, quality management, social responsibility, environmental responsibility, standards of compliance, certifications, and certifications with respect to specific vendor standards.
In the aspect, the entity score for a supplier is based on one or more of: loyalty as indicated by analysis of customs transactions, an amount of experience as indicated by customs transactions, evaluating the number of shipments, a duration of experience as indicated by shipments, size of transactions as indicated by past shipments, extent of international experience as indicated by past shipments, on extent of country-relevant experience as indicated by past shipments, and customer loyalty and supplier experience as indicated by past shipments reflected in customs records. In the aspect the entity score for a buyer is based on one or more of: loyalty as indicated by analysis of customs transactions, an amount of experience as indicated by customs transactions, evaluating the number of shipments, a duration of experience as indicated by shipments, a size of transactions as indicated by past shipments, an extent of international experience as indicated by past shipments, aggregated transactional customs data, customs data related to transactions of the buyer with a third party, and an extent of country-relevant experience as indicated by past shipments. In the aspect the entity score is for a plurality of factories of an entity. In the aspect, methods and systems further include providing a human-aided assessment of supplier skills as a factor in a entity score or include using an indicator of an entity's financial health as a factor in a entity score.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of records of customs transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; associating the transactions with entities; and determining a risk profile based on analysis of the aggregated transactions. In the aspect the risk profile is provided with respect to a supplier based on transactional customs data for the supplier. In the aspect, the risk is related to at least one of: counterfeiting, capacity, subcontracting, a political factor, a geographic factor, a weather factor, a geology factor, a financial risk, a probability of non-performance of a contract, a probability of termination of a contract, intellectual property, achieving a targeted delivery date. In the aspect the risk profile is provided with respect to a supplier based on transactional customs data for a party other than the supplier. In this aspect, the risk is related to at least one of: counterfeiting, capacity, subcontracting, a political factor, a geographic factor, a weather factor, a geology factor, a financial risk, a probability of non-performance of a contract, a probability of termination of a contract, intellectual property, achieving a targeted delivery date.
In the aspect, the risk profile is provided with respect to a buyer based on transactional customs data for the buyer. The risk is related to non-payment or the likelihood that a buyer will move to an alternative supplier.
In the aspect, the risk profile is provided with respect to a buyer based on transactional customs data for a party other than the buyer. The risk is related to non-payment or the likelihood that a buyer will move to an alternative supplier.
In the aspect, the risk profile is provided for a party using customs data and using the risk profile as a basis for determining terms and conditions of insurance.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of records of customs transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; associating the transactions with entities; and providing an indicator of economic leverage with respect to an entity based on analysis of the aggregated transactions. In the aspect, the indicator of economic leverage is with respect to at least one of: a supplier based on transactional customs data for the supplier, a supplier based on transactional customs data for a party other than the supplier, a buyer based on transactional customs data for the buyer, a buyer based on transactional customs data for a party other than the buyer. In the aspect, the transactional customs data corresponds to a price. Alternatively in the aspect, the transactional customs data corresponds to a delivery date or an order quantity.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include using a computer implemented facility to collect and store a plurality of records of customs transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; associating the transactions with entities; and predicting an action of an entity based on analysis of the aggregated transactions. In the aspect the prediction is of an action of a buyer based on analysis of customs data for transactions by the buyer. In the aspect the prediction is related to at least one of: price, a change in price, a change in supplier, and a quantity ordered by the buyer. In the aspect the prediction is of an action of a buyer based on analysis of customs data for transactions by a party other than the buyer. The prediction is related to a price, a change in price, a change in supplier, or a quantity ordered by a buyer. In the aspect, the prediction is of an action of a supplier based on analysis of customs data for transactions by the buyer. The prediction is related to a price, a change in price, a change in availability of an item, whether a supplier will work with a buyer of a given size, or whether a supplier will work with orders of a given size. In the aspect the prediction is of an action of a supplier based on analysis of customs data for transactions by a party other than the buyer. The prediction is related to a price, a change in price, a change in availability of an item, a potential closure of a subsidiary, a potential closure of a factory, or a potential closure of a company.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of records of customs transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; associating the transactions with entities; and making a recommendation based on analysis of the aggregated transactions. In the aspect the recommendation is based on analysis of customs data for transactions by the buyer, analysis of customs data for transactions by a party other than the buyer, analysis of customs data for transactions by the buyer, analysis of customs data for transactions by a party other than the buyer, prioritization of factors by a user, or a user-specified rating factor.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of records of transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; associating the transactions with entities; and associating an entity type with at least one of the entities. In the aspect, a data merging facility automatically merges records based on similarity of data elements with a customs record. The data elements correspond to a name of the entity or an address of the entity. Alternatively in the aspect, a data merging facility suggests an association between records and a single entity. The entity type is derived from one or more commodity fields in the transactions, and at least one of the commodity fields includes a harmonic tariff system code, a commodity type, or both. In the aspect, associating an entity type is based on an analysis of free text data in a plurality of data fields of the transactions. Associating an entity may alternatively be based on machine learning of entity types from customs transactional data records. The transactions may be customs transactions.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of records of customs transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; and processing the data to associate a plurality of transactions associated with a plurality of different entity names to a single entity based on analysis of customs record data for the transactions associated with the plurality of different entity names. In the aspect, the processing is based on, a name of the supplier, a name of the buyer, an order quantity, a billing amount, a location of the buyer, a location of the supplier, a delivery date, order data, at least one string associated with a supplier name, or at least one string associated with a buyer name. In the aspect the processing involves removing blank spaces from a supplier name field or removing blank spaces from a buyer name field. In the aspect, the transaction is associated with a region of interest, an industry, past shipment data, a country-relevant experience, a number of shipments, a material, a product category, a technique, a name of the entity, an order quantity, a billing address, a targeted delivery date, or a capacity of the supplier.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of public records of transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; associating the transactions with entities; and evaluating legitimacy of feedback about an entity based on analysis of whether the feedback is associated with a transaction reflected in public records. The evaluation of legitimacy of feedback associated with the supplier is based on validation by a third party. The evaluation of legitimacy of feedback associated with the buyer is based on validation by a third party.
In the aspect the transaction is associated with a name of the entity, an order quantity, a billing address, a targeted delivery date, a capacity of the supplier, transaction customs data, a region of interest, an industry, a past shipment, a country-relevant experience, a number of shipments, a material, a product category, or a technique.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of public records of transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; associating the transactions with entities; and providing a computer-implemented tool for suggesting a marketing strategy for a supplier based on analysis of transactional data from the public records. In the aspect the transactional data is associated with a supplier, a buyer, region of interest, customs data, past shipment, country relevant experience, a number of shipments, a product category, a material, or a technique. The analysis of transactional data includes analysis of pricing, buyer behavior, or transactional data associated with a competitor of the supplier.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of records of transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions;
associating the transactions with entities; and providing a computer-implemented tool for suggesting a marketing strategy for a buyer based on analysis of the transactional data from the records. In the aspect, the transactional data is associated with a supplier, a buyer, a region of interest, customs data, a past shipment, a country relevant experience, a number of shipments, a product category, a material, or a technique. The analysis of transactional data includes analysis of pricing, buyer behavior, or analysis of transactional data associated with a competitor of the buyer.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of public records of transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; associating the transactions with entities; and providing a user interface whereby a user may search for at least one of a supplier and a buyer and retrieve relevant information based on the aggregated transactions data. In the aspect, the interface allows a tuple-based search. The tuple-based search relates to a capability with respect to at least one of a product, a material and a technique. In the aspect, search results are ranked based on a supplier rating.
In the aspect, the rating is based upon a country context of a party, business legitimacy of a party, whether a party is registered with government authorities, an assessment of a trading environment in a country, macroeconomic information, public recognition of a party, industry awards, industry certifications, amount of experience, number of shipments, duration of experience, size of transactions, extent of domestic experience, extent of international experience, caliber of customers, customer loyalty, degree of specialization, specialization in product categories, specialization in manufacturing techniques, specialization in materials, specialization in gender, feedback from customers, feedback from buyers, feedback on product quality, feedback on customer service, feedback on timeliness of delivery, feedback on language skills, feedback on sample making ability, respect for intellectual property, quality management, social responsibility, environmental responsibility, standards of compliance, certifications, or certifications with respect to specific vendor standards.
In the aspect, the search results are based on a risk profile.
In the aspect, the risk is related to counterfeiting, capacity, subcontracting, a political factor, a geographic factor, a weather factor, a geology factor, a financial risk, a probability of non-performance of a contract, a probability of termination of a contract, intellectual property, achieving a targeted delivery date.
In the aspect, the risk profile is provided with respect to a supplier based on transactional customs data for a party other than the supplier. The risk is related to counterfeiting, capacity, subcontracting, a political factor, a geographic factor, a weather factor, a geology factor, a financial risk, a probability of non-performance of a contract, a probability of termination of a contract, intellectual property, achieving a targeted delivery date, a buyer based on transactional customs data for the buyer, non-payment, or the likelihood that a buyer will move to an alternative supplier. In the aspect, the risk profile is provided with respect to a buyer based on transactional customs data for a party other than the buyer. The risk is related to non-payment or the likelihood that a buyer will move to an alternative supplier.
In the aspect, the risk profile is provided for a party using customs data and using the risk profile as a basis for determining terms and conditions of insurance. The results are based on an opportunity profile. The opportunity relates to the availability of pricing leverage for a buyer with respect to a supplier, consolidation of orders with a supplier. The opportunity relates to the availability of pricing leverage for a supplier with respect to a buyer or to increasing a share of a buyer's total spending for a supplier.
In another aspect of the invention, the methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of public records of transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; associating the transactions with entities; integrating the aggregated and associated transactions with data from at least on 3 other data source to provide an integrated data facility; and adapting the integrated data facility for evaluating at least one of a supplier and a buyer. In the aspect, the public records include customs records. In the aspect, evaluations are ranked based on a supplier rating. In the aspect, the evaluation is based upon a country context of a party, business legitimacy of a party, whether a party is registered with government authorities, an assessment of a trading environment in a country, macroeconomic information, public recognition of a party, industry awards, industry certifications, amount of experience, number of shipments, duration of experience, size of transactions, extent of domestic experience, extent of international experience, caliber of customers, customer loyalty, degree of specialization, specialization in product categories, specialization in manufacturing techniques, specialization in materials, specialization in gender, feedback from customers, feedback from buyers, feedback on product quality, feedback on customer service, feedback on timeliness of delivery, feedback on language skills, feedback on sample making ability, respect for intellectual property, quality management, social responsibility, environmental responsibility, standards of compliance, certifications, or certifications with respect to specific vendor standards.
In the aspect, evaluations are ranked based on a buyer rating, a country context of a party, business legitimacy of a party, whether a party is registered with government authorities, an assessment of a trading environment in a country, macroeconomic information, public recognition of a party, industry awards, industry certifications, amount of experience, number of shipments, duration of experience, size of transactions, extent of domestic experience, extent of international experience, caliber of customers, customer loyalty, degree of specialization, specialization in product categories, specialization in manufacturing techniques, specialization in materials, specialization in gender, feedback from customers, feedback from buyers, feedback on product quality, feedback on customer service, feedback on timeliness of delivery, feedback on language skills, feedback on sample making ability, respect for intellectual property, quality management, social responsibility, environmental responsibility, standards of compliance, certifications, or certifications with respect to specific vendor standards.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of records of customs transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; associating the transactions with entities; and suggesting an opportunity based on analysis of the transactions. In the aspect, the opportunity relates to the availability of pricing leverage for a buyer with respect to a supplier, an opportunity for consolidation of orders with a supplier, the availability of pricing leverage for a supplier with respect to a buyer, the opportunity to increase a share of a buyer's total spending for a supplier, the availability of a discount for the buyer with respect to the supplier for a specified period, the availability of a committed time for delivery by the buyer to the supplier, the availability of bulk discount for the buyer with respect to the supplier, the availability of credit sales for the buyer with respect to the supplier, the availability of free delivery for the buyer with respect to the supplier, or the availability of liquidated damages for the buyer with respect to the supplier.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of aggregated customs transactions; associating the transactions with a supplier; and using the aggregated transactions to inform a rating of the supplier based at least in part on analysis of the aggregated transactions. In the aspect, the aggregated customs transactions include a summary of transactions for a product type. The transactions are summarized over a period of time. The analysis of the aggregated transactions includes comparing the aggregated transactions for a supplier with a plurality of records of transactions for a buyer. The aggregated customs transactions include transactions for a plurality of suppliers. In the aspect, associating the transactions with a supplier includes predicting one or more suppliers to which the transactions can be associated.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: taking a plurality of input data records from at least one data source of transactions; matching the data records to an entity that is a party to a plurality of transactions; and automatically merging the data records associated with the same entity to form a merged data store of transactions. In the aspect, matching includes filtering the data records. Filtering suggests data records for merging. Filtering is based on search engine techniques, such as a lucene search engine technique. Filtering is based on kgram filtering that may include a kgram filtering group that consists of four consecutive characters. In the aspect, a kgram filtering threshold for suggesting data records to be merged is ten matching kgram filter groups. In the aspect, a plurality of data fields within a data record are combined for matching.
In the aspect, matching includes classification. Classification is performed on data records suggested for merging and optionally, filtering is used to suggest data records for merging. Classification includes at least one of canonical adaptation, text cleanup, multi-field classification, edit distance assessment, vector generation, machine learning, and decision tree processing. Canonical adaptation includes normalizing text strings among the plurality of transactions or changing equivalent text strings to a known text string. Text cleanup is based on at least one of geographic factors, regional factors, market verticals, industry norms, known variations, learned variations, and user preferences.
In the aspect, the text cleanups are associated with at least one type of data field in the data records. The type of data field includes at least one of a shipper, a consignee, a notify party, an also notify party, a weight, a quantity, a country, a date, a commodity, and a harmonic tariff system code. Classification is applied to a plurality of data fields in the data records or to combined data fields in the data records.
In the aspect, classification provides a vector that represents dimensions of similarity. The vector includes dimensions of similarity for at least two of canonical adaptation, text cleanup, multi-field classification, edit distance assessment, vector generation, machine learning, and decision tree processing.
In the aspect, matching includes clustering. Optionally, clustering includes p-percent clustering. In the aspect, a data record is merged when a p-percent value associated with the data record exceeds a p-percent threshold associated with the entity. Optionally the p-percent threshold is thirty percent. Alternatively p-percent clustering is based on a dynamic p-percent threshold. The dynamic p-percent threshold is based on a quantity of data records in a cluster associated with an entity.
In the aspect, the plurality of transactions contains party identifying data in a field of the data records. The party identifying data is stored in different fields of at least two of the plurality of data records. Optionally, party identifying data in a first record is a parent entity and a party identifying data in a second record is a child entity of the parent entity.
In the aspect, matching data records includes identifying data that is a variation of an entity name or an entity address. In the aspect, the party is one of a supplier and a buyer. Alternatively, matching includes two or more types of text association selected from a list consisting of: filtering, character group matching, thesaurus lookup, machine learning, natural language processing, search-based comparison, classification, known entity matching, clustering, and human-identified entities.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: taking a plurality of input data records from at least one data source of transactions; filtering the input data records to identify a set of filtered data records that are favorable candidates for automatic merging; classifying the filtered data records to produce a set of classified data records, each classified data record associated with a likelihood that the data record should be associated with a particular entity; and automatically merging the data records associated with the same entity to form a merged data store of transactions. In the aspect, the filtering is performed using a search engine, kgram filtering, or dynamic programming. In the aspect, classifying the data records is performed using at least one of canonical adaptation, specific cleanups, multi-field comparison, an edit distance algorithm, vector generation, machine learning, and a decision tree. In the aspect, filtering suggests data records for merging. Alternatively filtering is based on search engine techniques, that optionally include a lucene search engine technique. In the aspect filtering is based on kgram filtering. Optionally a kgram filtering group consists of four consecutive characters. Optionally a kgram filtering threshold for suggesting data records to be merged is ten matching kgram filter groups. In the aspect, a plurality of data fields within a data record is combined for matching.
In the aspect, classification is performed on data records suggested for merging. Optionally filtering is used to suggest data records for merging. In the aspect classification includes at least one of canonical adaptation, text cleanup, multi-field classification, edit distance assessment, vector generation, machine learning, and decision tree processing. Canonical adaptation includes normalizing text strings among the transactions or changing equivalent text strings to a known text string.
In the aspect, text cleanup may be based on at least one of geographic factors, regional factors, market verticals, industry norms, known variations, learned variations, and user preferences. Optionally, text cleanups are associated with at least one type of data field in the data records. The type of data field includes a shipper, a consignee, a notify party, an also notify party, a weight, a quantity, a country, a date, a commodity, and a harmonic tariff system code.
In the aspect, classification is applied to a plurality of data fields in the data records or to combined data fields in the data records.
In the aspect, classification provides a vector that represents dimensions of similarity. Optionally, the vector includes dimensions of similarity for at least two of canonical adaptation, text cleanup, multi-field classification, edit distance assessment, vector generation, machine learning, and decision tree processing.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of records of transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; associating the transactions with entities; and classifying an entity as a buyer based on analysis of the aggregated transactions. In the aspect the aggregated transaction is associated with an industry, customs data, a past shipment, a likelihood of interest, or a number of shipments.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of records of transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; associating the transactions with entities; and using the transactions as a training set to predict association of a particular transaction with an attribute. In the aspect, the attribute is a type of industry, a type of supplier, a type of product, a product attribute, or related to a type of material. In the aspect, the particular transaction represents a shipment from a supplier to a buyer. The transactions are customs transactions. Alternatively, the entities are one or more of a supplier and a buyer. Optionally, the particular transaction is a rolled-up transaction.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of records of transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; associating the transactions with entities; and using the transactions as a training set to predict association of a particular transaction with an entity. In the aspect, the particular transaction represents a shipment from a supplier to a buyer. Alternatively, the transactions are customs transactions. In the aspect, an entity is one of a supplier and a buyer. Optionally, the particular transaction is a rolled-up transaction.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of records of transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; associating the transactions with entities; and predicting a minimum order requirement for an entity based on analysis of the transactions. In the aspect the entity is a factory, supplier, or a subsidiary of a supplier.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of records of transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; associating the transactions with entities; and providing a search facility for enabling a search for an entity, wherein the search facility allows searching based on geographic region, industry specialization, entities participating in the transactions, and likelihood of interest in a transaction with the searcher. In the aspect, the search facility is adapted to be used by a buyer searching for a supplier or adapted to be used by a supplier searching for a buyer.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of records of transactions among a plurality of buyers and a plurality of suppliers; aggregating the transactions; associating the transactions with entities; and rating a sub-entity of a supplier based on analysis of the aggregated transactions. In the aspect, the sub-entity is a factory, a collection of factories, or a subsidiary. In the aspect, determining the sub-entity is based on analysis of the public records. Optionally, the public records are records of customs transactions.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of aggregated public records of shipment transactions; associating the transactions with a supplier; and using the aggregated transactions to inform a rating of the supplier based at least in part on analysis of the aggregated transactions.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of public records of transactions; associating the transactions with entities; and using the aggregated transactions to classify at least one of a supplier and a buyer according to type. In the aspect a buyer may identify like buyers, suppliers like those of the buyer, suppliers of a specified type, or suppliers like to prefer the buyer. In the aspect, a supplier may identify like suppliers, buyers like those of the supplier, or buyers of a specific type.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of public records of transactions; associating the transactions with entities; and assessing whether a buyer has ceased doing business with a supplier based on the transactional data. In the aspect, the assessment is based on cycle time between shipments, departure of cycle time from a historical average, or based in part on a prediction as to inventory held by a buyer.
In another aspect of the invention, methods and systems, such as computer implemented methods and systems, include: using a computer implemented facility to collect and store a plurality of public records of transactions; associating the transactions with entities; and using the aggregated transactions identify at least one of a supplier of a specific item sold by a party other than the supplier. In the aspect, the specific item is a commodity.
In the aspect, the identification of supplier is based on region of interest, customs data, product category, past shipments, or a number of shipments. The specific item is a service.
These and other systems, methods, objects, features, and advantages of the present invention will be apparent to those skilled in the art from the following detailed description of the preferred embodiment and the drawings. All documents mentioned herein are hereby incorporated in their entirety by reference.
The invention and the following detailed description of certain embodiments thereof may be understood by reference to the following figures:
Methods and systems are provided herein for facilitating engagement of suppliers; thus, a supplier rating facility may make it easier for companies of all sizes to do business across borders by helping companies identify which suppliers they can trust. The suppler rating facility approach is to leverage a wide variety of quality data sources to rate suppliers around the globe. Behind each rating may be a detailed scorecard that evaluates suppliers along key dimensions. By comparing supplier scorecards, subscribers may determine which suppliers are right for them. In one preferred embodiment the rating system is used to rate apparel suppliers, but it should be understood that suppliers in other industries may be rated by the same or similar methods and systems, such as suppliers of consumer electronics, computer equipment, toys and games, consumer products, textiles, home goods, food, accessories, computer games, automotive parts, electronic parts and equipment, and a wide range of other goods and services, such as BPO, software development, call centers, and the like.
Presently, buyers can access a plurality of supplier directories for information about suppliers. However, those directories may only contain information provided by the suppliers themselves, and on occasion, third-party information on limited subjects, such as relating to creditworthiness. That information may not be particularly useful in helping customers to distinguish between good and bad suppliers. In certain preferred embodiments, the supplier rating facility disclosed herein may facilitate the generation of a plurality of reports that supplement or substitute for supplier-provided information, the reports generated by methods and systems disclosed herein and based on a wide range of data sources. In embodiments each supplier may receive a rating between 1 and 100. Behind this rating may be a detailed scorecard, each component of which being generated by an algorithm that operates on one or more relevant data sources, and that evaluates suppliers along dimensions that are important to customers.
A supplier rating facility as contemplated herein may provide buyers with concrete information about which suppliers are good and which suppliers are bad, which are trustworthy and which are not, which are experienced in a particular area, and the like. The ratings may feature a range of information about suppliers, including analysis generated by algorithms operating on relevant data sources and, in certain optional embodiments, ratings from previous customers. Analysis may include, among other things, using publicly available but currently fragmented information. In various embodiments, the supplier rating facility may rate suppliers along several dimensions, including without limitation amount of international experience, degree of specialization, and standards compliance.
In certain optional embodiments, ratings from previous customers may enable suppliers to gather and showcase feedback from their previous customers. Buyers may pay a subscription fee for access to ratings detail. Existing business-to-business sites may be able to embed the supplier rating facility in their directories and benefit from new revenue streams. Although apparel is being used as an embodiment of the invention, it should be understood that the invention may be applied to any industry, such as furniture, electronics, textiles, chemicals, toys, food, and the like. In addition, services in addition to a ratings service may be facilitated through the invention, such as for billing, transactional settlement, insurance, social networking for buyers, and the like. In embodiments, the invention may be applied to a broad spectrum of industries where buyers and sellers are located across diverse environments, and supplier-product information and ratings are fragmented.
The supplier rating facility may provide a ratings platform where buyers/suppliers may compare and contrast potential suppliers/buyers. The ratings platform may generate and maintain ratings of suppliers, buyers, countries, geographic regions, marketplaces, commodities and the like. The ratings may be presented in various forms including a listing of supplier ratings as shown in
The list 100 may include entries 108 for each supplier that satisfies the keyword 102 criteria. An entry 108 may include an overall rating 110 also known as the “Panjiva Rating”, the supplier name 112, selected bibliographic data 114, and the like. Preferences as indicated above may impact what information is presented in an entry 108 and the embodiment of
An alternate view of supplier rating is exemplified in
In embodiments, a backend infrastructure may automatically generate customized documents by programmatically generating a representation of the document in a typesetting language such as Tex, LaTeX, and the like, which may then be processed and turned into a PDF document.
The “Business Basics” section of the Supplier Scorecard 200 may help a buyer assess whether a company is legitimate and worthy of consideration as a potential partner. Included in “Business Basics” may be information on whether a company has registered with authorities, as well as an assessment of a trading environment in the supplier's country, combining macro contextual information with data that is specific to an individual supplier and the like. A facility for determining a track record in a particular jurisdiction may use government and third-party data, and may assess the amount of experience a supplier has serving that jurisdiction, and the loyalty that a supplier's customers have demonstrated, and the like. The “Standards Compliance” section of the Supplier Scorecard 200 may document whether a supplier has been certified as meeting international standards for quality management, respect for the environment, social responsibility, product safety, and the like.
The ratings scorecard 200 may include a plurality of analysis dimensions, such as county context, business legitimacy, public recognition, amount of experience, caliber of customers, customer loyalty, specialization, quality management, social responsibility, environmental responsibility, and the like. Buyer feedback dimensions may include product quality, customer service, timeliness of delivery, language skills, sample-making ability, respect for intellectual property, and the like. Supplier information may include contact information, areas of expertise, caliber of customers, ratings, and the like. A confidence in buyer feedback may be established by determining that the feedback is being provided by a supplier who is or recently was receiving shipments from the supplier. This can be done by ensuring that transaction records validate that at least some supplier shipments were supplied to the buyer providing the feedback. In embodiments, information utilized in the formation of the ratings scorecard 200 may be from shipment history, such as frequency, quantity, and the like; shipment capacity estimation, which may be based on shipment data as opposed to information provided by the supplier.
Contact information may include making all contact information available to subscribers, so that they may directly contact suppliers. Areas of expertise may tell a buyer which products a supplier has shipped, which materials it has used, which techniques it has employed, and whether it has produced men's apparel, women's apparel, or both. Caliber of customers may tell a buyer which types of customers a supplier has served, such as premium, mass, discount, and/or niche customers. In embodiments buyers may rate suppliers with whom they have done business. After a buyer rates a supplier, the supplier rating facility may verify that that the two have actually done business together, such as by identifying a corresponding customs records that shows an actual import transaction in which the buyer imported goods from the supplier, from a bill of lading, from a bank-issued receipt, and the like. Thus, methods and systems disclosed herein include methods and systems for deterring fraudulent ratings by verifying the existence of the transaction purportedly rated by the buyer. This may prevent false ratings that are either too positive (such as by an affiliate or cohort of the supplier) or too negative (such as by a competing supplier posing as a buyer). After verification, the buyer's rating may become part of the supplier's scorecard. As part of the verification process, the buyer's identity may be revealed to the supplier. However, in embodiments the buyer's identity may be obscured so that it does not appear on the supplier rating facility's website and is not shared with anyone else. In embodiments buyer feedback may only be viewed by buyers who have provided feedback on their suppliers. In embodiments, a computer facility for recording transactions associated with one or more buyers with one or more sellers may include a user interface that may facilitate determination of entity score based on transactional data. The transactional data may be related to the shipping details of the goods and services associated with different entities. In an example, entities such as buyers may order goods and services from the sellers resulting in transactions. An aggregation facility may collect, combine or aggregate transactions associated with different entities. Subsequently, an association facility may facilitate association of transactions with different entities. The transactions may be analyzed by the analysis facility to generate an entity score corresponding to each entity.
In embodiments, rating a supplier, buyer, or other entity may result in a score that is at least partially based on predefined criteria, such as a user provided criteria. Alternatively, the system and methods herein may facilitate rating of a supplier, buyer, or other entity based one or more algorithms. The rating algorithms may be manually selected, or may be selected automatically based on a set of algorithm selection rules. In an example, a supplier may be known in the industry as highly credible. One or more rating algorithms may be applied to transaction data and may use predefined criteria for the algorithms to mathematically determine the credibility of the supplier. This determined credibility rating may be provided to the buyer through a user interface of the platform.
In embodiments, the entity score may be based in part on transactional data related to the shipments by the entity such as delivery data, amount shipped, location of shipment and the like. In an example, a supplier providing goods and services within the stipulated delivery date may garner higher ratings compared to the suppliers who failed to deliver on time.
In embodiments, the entity score may be based on one more factors including country context, business legitimacy information, public recognition, amount of experience, caliber of customers of the supplier, customer loyalty for the supplier, degree of specialization of the supplier, and feedback from previous customers or some other factors. Further, each factor or group of factors may include a list of parameters. A user interface may be configured to allow a user to select some or all the parameters from this group to generate an entity rating. In an example, the group country content may include variables such as GNI per capita, currency volatility, cost to export, political stability, and the like. The user interface may allow a user to select GNI per capita and cost to export to generate a country context value that would be applied to calculate an entity rating. Furthermore, determination of the entity score may depend partially or completely on some or all the parameters selected from some or all the groups, as described herein and elsewhere. In an example, a buyer who may be interested in knowing the quality of a product or service provided by a supplier may select feedback from previous customer groups on which to base an entity score for the supplier. This group may further include parameters such as timely delivery of goods, quality of goods, number of transactions and the like. Rather than choosing just the group rating to determine an entity rating, the buyer may choose some or all the parameters from this group to determine the entity score associated with the supplier. In another example, the score associated with the supplier may be determined based on two or more groups comprising multiple parameters, such as the group pertaining to the degree of specialization of the supplier and the group pertaining to feedback from previous customers. The buyer may select the group degree of specialization and one or more parameters from the group. Similarly, the buyer may select one or more parameters from the feedback from the previous customers group. The entity score may be determined based on parameters selected in each of the groups.
A user interface, such as the user interface of
In embodiments, a supplier rating facility or buyer rating facility may take ratings along each of key dimensions, weight the ratings to account for the fact that some dimensions are more important than others, calculate an overall rating 110, and the like. In embodiments, ratings may provide a measure of caliber, such as the caliber of a buyer or the caliber of a supplier.
The supplier rating facility may rate suppliers across a plurality of different dimensions, some of which may derive from actual transactional data, such as customs data, with others based on sources such as Dun & Bradstreet, the World Bank, auditing firms for various certifications, government sources, and the like. In embodiments, more weight may be given to recent data, data for larger transactions, data for higher quality buyers, or other types of data with respect to which there is an indicator that the data may have higher relevance than other types of data. The supplier rating facility may also provide a more intuitive understanding to ratings, by considering caliber of customers, customer loyalty, specialization, and the like. Caliber of customers may involve manually grouping buyers into distinct bands or tiers, such as premium, mass-market, discount, niche, and the like and then computing a sum based on the newness of each buyer-supplier relationship and the tier of each buyer.
In embodiments, rating for a supplier may be based on the aggregated transactional customs data, a user defined criteria, customs data related to transactions by the supplier with a third party or on some other parameters. In an example, a supplier may be rated based on the number of transactions done with a particular buyer. In another example, a user may define adherence to delivery data as a criteria for rating the supplier. In addition, the rating of the supplier may be based at least in part on loyalty as indicated by an analysis of customs transactions. Furthermore, the determination of rating for the supplier may based on an amount of supplier experience as indicated by customs transactions related to the number of shipments, duration of supplier experience as indicated by shipments, size of transactions as indicated by past shipments, extent of international experience as indicated by past shipments, extent of country-relevant experience as indicated by past shipments and the like.
In embodiments, a rating for a buyer may be based on aggregated transactional customs data, customs data related to transactions of the buyer with a third party or some other parameter. In an example, ratings for the buyer may be based on the feedback about the buyer provided by one or more suppliers. In addition, ratings for the buyer may also be based on two or more factors selected from a group including the country context of a party, the business legitimacy of a party, whether a party is registered with government authorities, an assessment of a trading environment in a country, macroeconomic information, public recognition of a party, industry awards, industry certifications, amount of experience, number of shipments, duration of experience, size of transactions, extent of domestic experience, extent of international experience, caliber of customers, customer loyalty, degree of specialization, specialization in product categories, specialization in manufacturing techniques, specialization in materials, specialization in gender, feedback from customers, feedback from buyers, feedback on product quality, feedback on customer service, feedback on timeliness of delivery, feedback on language skills, feedback on sample making ability, respect for intellectual property, quality management, social responsibility, environmental responsibility, standards of compliance, certifications, and certifications with respect to specific vendor standards and the like.
In embodiments, the rating of the buyer may be based on loyalty as indicated by an analysis of customs transactions. In an example, a buyer may be rated based on the number of transactions with a particular supplier in a specific time frame. In addition, the buyer may be rated based on an amount of experience as indicated by customs transactions related to the number of shipments, duration of experience as indicated by shipments, size of transactions as indicated by past shipments, extent of international experience as indicated by past shipments, extent of country-relevant experience as indicated by past shipments and the like.
In embodiments, a rating may be made of customer loyalty for a supplier. A customer loyalty rating method may include analyzing the set of buyers who have done business with each supplier over the course of several years, and identifying ‘loyalty periods,’ intervals in which a buyer consistently sources from a given supplier, and ‘switches,’ where a buyer ceases obtaining a given set of products from one supplier and begins sourcing from another supplier. Suppliers for whom there have been many switches may be given lower ratings, while suppliers with long loyalty periods and few switches may be given higher ratings.
In embodiments, a rating may relate to the degree of specialization of a supplier. A specialization rating method may decompose supplier shipments into dimensions, such as product category, technique, material, gender, and the like. These dimensions may be independent of the rating dimensions or may be used as a factor in rating.
In embodiments, methods and systems may include methods for generating raw scores. Generation of raw scores may use a variety of techniques to transform raw customs data and other third-party data into meaningful ratings. Consideration may be given to customer loyalty, caliber of customers, amount of experience, specialization, country context, business legitimacy, environmental responsibility, social responsibility, quality management, public recognition, and the like. Customer loyalty rating may include identifying shipping patterns, buyer patterns, loyalty periods, and the like. Caliber of customer ratings may include assigning a buyer tier, length of time in that tier, age of buyer, and the like. Experience ratings may include evaluating number of shipments, duration of experience, size of transactions handled, and the like. Specialization ratings may include or reference a measure of the extent to which a supplier focuses on a narrow range of products, materials, and/or techniques. Business legitimacy ratings may be provided by a supplier having government registration records, a Dun & Bradstreet DUNS number, or other evidence of business legitimacy. Environmental, social, product safety, and quality management ratings may be derived from a supplier having appropriate certifications, and the like. Public recognition ratings may include reference to government and industry awards, and the like. In embodiments, high risk suppliers and high risk buyers may be identified, such as in association with individuals and organizations that work with high risk suppliers and high risk buyers. Country context ratings may be related to the country in which a supplier is located, as well as data supplied by the World Bank, International Monetary Fund, and other sources, about that country. Other sources may include GNI per capita, currency volatility, cost to export, political stability, credit_rank, export_cost, gci_efficiency_enhancers, and the like. A country context computation may include calculating a weighted sum of log(gni_per_capita), credit_rank, log(export_cost) and gci_efficiency_enhancers that may be then thresholded into final score buckets. The weights and thresholds used in the country context computation may have been determined using machine learning techniques (e.g. decision trees and principal component analysis) to determine the relevant features to weight, appropriate weights, and effective thresholds.
Generating ratings from raw scores may include a weighting, standardization or normalization factor applied to raw data to produce a standard score that may be centered on zero, and then normalized to a rating between 0 and 100. These values may then be applied to a scorecard 200 that presents the normalized data for a supplier. In embodiments, the ratings may be scaled linearly to provide a mean of approximately 50, as in a Gaussian distribution.
In addition, ratings may be customized to individual buyer preferences, such as by having buyer's rate suppliers with whom they have done business. Ratings may then be tuned to best match this empirical view of a buyer's preferences. Such an approach may use a machine learning technique such as a support vector machine. Over time, trends in ratings may then be captured and displayed to the buyer. Such trends may enable a graph-theory analysis (e.g., minimum cut, maximum flow, cliques, and the like) on buyer-supplier networks to determine the relationships between groups of buyers and suppliers, which may lead to additional value-added services such as improving production allocation for buyers.
Referring to
In embodiments, public records may include customs records apart from other records. The customs record may include information associated with an entity as captured by a customs organization. The information may be useful in identifying the different transactions associated with the entity's billing based on the entity's custom identification number.
Data sources may leverage data from several hundred data sources, such as, International Oeko-Tex Association, Social Accountability International, Worldwide Responsible Apparel Production (WRAP), Fifty-five ISO 9001 auditing firms, Forty-six ISO 14001 auditing firms, Forty-seven OHSAS 18001 auditing firms, Two GB/T 18885 auditing firms, United States Department of Homeland Security, Ministry of Commerce of the People's Republic of China, General Administration of Customs of the People's Republic of China, and the like. In embodiments, custom data may be from countries all over the world, covering exports and imports, where an import record may be matched up with an export record.
Tools used in the analysis of supplier and buyer data may include a merger tool, a suggestive merger tool, a buyer caliber tool, a buyer marketing tool, a name chooser tool, a country manager, an API, a product keyword manager tool, a statistics tool, a report generation tool, a supplier marketing tool, a name updater, and the like. Any of these tools may be embodied in the facilities of computer facility 302.
In embodiments, aggregated customs data may be processed to identify transactions associated with different types of entities such as buyers and/or suppliers. In addition, based on the transactions associated with different entities, an entity type may be determined from one or more entity types present in the transactions. In an example, an entity may supply woolen clothes and the transactions associated with the shipments of the woolen clothes may be recorded as being provided by ‘ABC co’. In another transaction, the same entity may be recorded as ‘ABC Company’. This variation may be due to a difference in recording of transactions of customs data due to variations in filling data in the customs form rather than the entities being different entities. A data merging facility may allow automatic merging of transaction records described above under a single entity based on the inference made on grounds of similarity of data. In this example, the variations ‘ABC co’ and ‘ABC Company’ may form a valid case of merging of data based on the minor variation in entity name. Alternatively to automatic merging, or in addition to it, suggestions for merging similar data based on similarity in data elements may be provided to a user. In the above example, records relating to transactions ‘ABC co’ and ‘ABC Company’ may be presented to the user with a suggestion to merge them under a single entity based on similarity of data elements. The similarity of data element in the records may be determined by the data merging facility. Based on the user's response, automatic merging of the two entity names may be learned by the platform.
An entity may be associated with one or more names for performing transactions that may be captured in customs records. As described above, the difference may be due to a variation in recording of the customs data. The systems and methods described herein may facilitate merging of any number of transactions that should be associated with a particular entity even though the records show a plurality of similar but varied entity names.
In addition to facilitating processing aggregated customs data so as to associate a set of transactions with an entity, a plurality of transactions that are properly associated with a plurality of entities may be merged under an entity type for purposes of evaluating the transactions and entities associated therewith. The merged records may be useful in evaluating a market segment, consortium of companies, industry segment, regional results, class of entities, and the like. In an example, transactions associated with several entities may be merged based on the basis of the transactions being associated with a single buyer. Even though the transactions call out different suppliers in different industries, the single buyer is a basis for processing the transactions as if they were merged.
An entity type may be defined based on any aspect of an entity that may be used to process customs transaction records. The methods and systems of filtering, classification, and clustering of transactions as described herein may be applied to identify transactions that are mergeable under an entity type. In an example, a buyer may initiate purchase transactions with four suppliers of components to produce an item. The transactions between any of the four suppliers and the common buyer can be merged (or tagged as mergeable) as having a common entity type such as “supplier to common buyer”. Other suppliers who ship items to the common buyer may have their transactions with the common buyer merged under the same entity type.
In many cases multiple data records exist for a single supplier, but the relationship of those records to that single supplier is ambiguous. In an example, the name of the supplier might appear in one field in one record, but in a completely different field in another record. This is often the case in customs data, where forms are filled out with information in various fields, notwithstanding the purported standardization of the forms. In embodiments, a merger tool may be used to merge data records of two apparent suppliers that should really be one supplier. The merger tool may evaluate an address, and if the same in two records, select a parent record and identify child records, upon which records are merged into a single record. In embodiments, a merger tool may merge records that are on the same page or more generally merge records across a database. In embodiments, a merger tool may use a pattern matching technique to identify potential candidates for merging of records
As shown in
In embodiments, the indicator of economic leverage may be with respect to the supplier 428 based on transactional customs data for the supplier 428. In embodiments, the indicator of economic leverage may be with respect to a supplier 428 based on transactional customs data for a party other than the supplier 428. In embodiments, the indicator of economic leverage may be with respect to a buyer 420 based on transactional customs data for the buyer 420. In embodiments, the indicator 420 of economic leverage may be with respect to a buyer 420 based on transactional customs data for a party other than the buyer 420.
In embodiments, as shown in
In embodiments, the prediction may be of an action of the buyer 420 based on an analysis of customs data for the buyer 420 transactions. The prediction may be related to a price, a change in price, a change in supplier, a quantity ordered by the buyer 420 and the like. In embodiments, the prediction may be of a buyer action of based on an analysis of customs data for transactions by a party other than the buyer 420. In embodiments, the prediction may be of a supplier action based on an analysis of customs data for transactions by the buyer 420. In embodiments, the prediction may be of a supplier action based on an analysis of customs data for transactions by a party other than the buyer 420. In embodiments, the prediction may be related to a potential closure. The closure may be of a subsidiary, a factory, a company and the like. Those skilled in the art would appreciate that the prediction facility 502 may provide the predictions to the plurality of buyers 418, plurality of suppliers 424 or some other entities.
In embodiments, as shown in
In embodiments, the recommendation may be based on analysis of customs data for the buyer 420 transactions. In embodiments, the recommendation may be based on analysis of customs data for transactions by a party other than the buyer 420. In embodiments, the recommendation may be based on analysis of customs data for transactions by the buyer 420. In embodiments, the recommendation may be based on analysis of customs data for transactions by a party other than the buyer 420.
In embodiments, the recommendation may be based on prioritization of factors by a user. In an example, the buyer 420 may require 40 tons of silk within 4 days. The recommendation facility 602 may recommend buying 40 tons of silk from supplier 430 based on its manufacturing capacity of 50 tons per day and ability to provide the required silk within the stipulated time. In embodiments, the recommendation may be based on a user-specified rating factor.
In embodiments, a suggestive merger tool may use more sophisticated techniques to suggest which buyers or suppliers should be merged together, such as the supplier in question listed, then producing potential matches. Such techniques may use text similarity metrics on the name and the address and performing algorithmic steps such as sorting the tokens by alphabetical order, so word transpositions do not change lexical distance of names in pattern matching. Such techniques may determine how each word in a given buyer or supplier's name or buyer's name contributes to the uniqueness of the name, and uses this information to make relevant suggestions for merging. In embodiments, a suggestive merger tool may use a machine learning approach to perform pattern matching or otherwise suggest merger of records, such as a technique with boosted trees or other machine learning techniques.
A buyer caliber rating may be assigned manually by checking a box based on a facilitator's or host's assessment of the caliber of the buyer, or by automated techniques. In embodiments, a search link may be provided for each buyer, such as one that retrieves search results from a search engine, directory, rating system or other source of information about the buyer. In embodiments, an interface, such as an overall buyer manager, may assist suppliers in searching for different buyers.
In embodiments, a buyer marketing tool may break down data for a particular supplier, such as addresses (from customs data), raw customs records, records that show customer loyalty periods and switches to other suppliers, specific breakdowns of what the supplier has shipped (e.g. in terms of product category, material, technique, gender of the shipped garments, and the like), breakdowns of the size of the shipments the supplier has made, breakdowns of the number of shipments that the supplier has made each month over some time period, as to determine the estimated capacity of a supplier, and the estimated minimum shipment that a supplier is willing to produce. In an example, a tool may show a breakdown of suppliers (e.g. showing a number of suppliers, such as 35 suppliers in ratings for each), where it is possible to see a history of which suppliers buyers have used. This may allow marketers to evaluate their performance relative to other suppliers with whom they compete.
A country manager tool may be used to add data on countries (such as for the country context dimension of an overall rating 110 or analysis).
In embodiments, an application programming interface may be provided for the platform described herein, whereby other computer programs may access the reports generated by the platform, such as accessing overall ratings 110, specific components of ratings, results of particular algorithms, or data sources used in the platform. Thus, other parties that engage in global trade, such as clients of the facilitator and partners, may obtain access to the platform, allowing the ratings managed by the platform to become a standard measure by which suppliers are rated.
A product keyword manager tool may provide an ontology or hierarchy for a search interface, such as using graphs, charts, and the like. The manager may allow a facilitator to add or delete sub-categories to a category. Keywords associated with each of the categories may be useful for: (1) getting the data (allowing a user to scan through raw text of customs data, looking for these keywords, which is one way for the facilitator to know that a supplier has shipped something within a category; In an example, a search for infant clothing might search for all sub-categories, using words such as baby, infant, kiddy, kid, layette, maternity, newborn, toddler, and the like); and (2) using keywords for text entry into the search field (such as synonyms to get better search results). In an example, in a hierarchy of materials, there are sub-materials of materials, and each has keywords associated with it. In embodiments, a facilitator may engage in a process (manual or automatic) to generate key words, such as using glossaries that list all of the products and materials, with specific definitions. In embodiments, algorithms may be used to determine the market vertical (e.g., apparel supplier or electronics supplier, etc.) of a supplier based on the aggregate contents of all of its shipments. In embodiments, customs records may be utilized to identify what industry or vertical the material is in.
A statistics tool may assist in providing distributions of data. Thus, a facilitator may support statistical distributions for all dimensions of data analyzed by the methods and systems of the platform.
A bucket boundary check tool may assist in testing that suppliers that fall within specific rating “buckets” or bins (e.g., Excellent, Unproven, etc.) by showing suppliers that are at the high and low boundary of each bucket.
A report generation tool may automatically build PDF reports or reports in other output formats, such as PowerPoint, Excel, Word, and the like. The report generation tool can be used as an administrative tool, or as a tool to allow users or clients to build custom reports, such as ones that incorporate some or all of the data generated by the platform. In an example, a user may specify which supplier or suppliers the report should include. A report could be made vertical-specific, covering a number of suppliers in a vertical, such as a theme or characteristic grouping, or it could relate to standout suppliers (such as ones with high customer loyalty ratings, top amounts of experience, or the like). In embodiments, a user may turn on or off various sections. In embodiments, the end product of a report may be a link that allows the user to download the report (in PDF format or some other format) or which sends the report via email (in PDF format or some other format).
In embodiments, a buyer marketing tool may provide information about what product materials, product techniques, and the like particular buyers require from their suppliers. Such a tool may also provide information relating to how many shipments particular buyers imported over time, as well as the breakdown of shipment sizes.
A marketing tool may be used by an operator of the platform to at least identify opportunities of marketing the products and services associated with the platform to suppliers, buyers, and others. The marketing tool can also be accessed by suppliers as shown in
The marketing tool may also work cooperatively with a user interface to facilitate an operator entering parameters of marketing opportunity scenarios that the marketing tool can evaluate. The entered scenario parameters and attributes may be applied to an analysis of customs transaction data and marketing opportunities may be presented to the operator through the user interface.
As shown in
As shown in
Processing customs transactions and other records may involve a multi-step method. Data from customs organizations, such as US Customs may be provided on a removable computer memory such as a CD, DVD, flash memory, memory stick, USB memory card, and any other type of removable or portable memory device. Alternatively the customs data may be acquired via a network, such as the Internet, a dialup connection, a virtual private network, a dedicated network, and the like. The data may also be converted from a proprietary format for further processing by the platform. Each customs organization, and within any customs organization for a particular country may have a different format or storage device for records. Conversion may be performed on the data so that the end result is independent of the physical format of delivery and the logical formatting of the information. In this way data in a substantially unified format may be processed by the methods and systems of supplier rating and the like that are herein disclosed. In an example, US customs data may be provided on a CD and may be in a COBOL format. The data on the CD may be retrieved and automatically loaded to a server. The server or another computing device may convert the data from the COBOL format to an XML format. The XML formatted data may be loaded to a database such as a Postgres database for further processing. In this example XML format represents a unified format for the customs data.
Processing the converted transaction data may include multiple steps of data analysis in which a confidence level may be applied. Confidence levels may be grouped into confidences bands that may help target each transaction toward one of merging (high confidence band), suggest human-aided merging (medium confidence band) and do not merge (low or lacking confidence band). Analysis of the transaction data may reveal important information about the entities involved in the transactions. In an example, a single entity may appear as a buyer in one transaction, a shipper in another and a supplier in a third. Ensuring that each transaction is properly associated with the entity as its intended function (buyer, shipper, and supplier) may be accomplished through various analysis and assessment techniques including, similarity assessment, filtering, classification, clustering, and the like.
Processing of customs data may also include text mining. Text mining may include searching for key words, terms, or phrases that are known, predetermined, or specified for the mining operation. An ontology of terms, such as ‘gender dyeing’ may be applied in text mining. In addition, synonyms of keywords may be mined. Text mining also facilitates populating reports with various data, such as time series data of shipments per month and weight of shipments per month.
Data may be further analyzed with a monitoring tool that may look for anomalies, such as peaks, and other statistical measures to identify potentially important events that are captured in the transactions. Analyzing data for peaks and the like may help with activating buyers, suppliers, shippers, and other entities for use in the platform. A statistical event, such as a spike in orders by a buyer that otherwise had little transaction history may trigger an indication that the buyer should be activated for use in ratings of buyers, suppliers, and the like. Alternatively, or in addition, an entity may be activated based on the transactions for the entity complying with criteria such as a shipment quantity threshold, and the like.
Merging data ensures that all records of transactions associated with an entity (buyer, supplier, or the like) are properly recorded against the correct entity. However, due to the large number of data sources, substantial variations in how an entity may be identified in records from the data sources, parent-subsidiary entity relationships, transaction system limitations (e.g. limiting the number of characters in an entity name), regional differences, dialect differences, use of short hand for entity information, various coding schemes used by buyers, suppliers, and the like proper merging of data is complex and difficult.
In a fundamental example, merging is taking two records for the same entity that each has entity information that varies substantially one from the other and ensuring that the records are properly recorded against the one entity rather than being assigned to two separate entities. A merger tool may provide robust, accurate, and efficient merging of data by resolving the variations, some of which are described above so that records for a single entity are merged, while ensuring that records for a different entity are kept separate from the single entity.
Within any given country, industry, region, or language there is no universal entity identifier that could be applied to the data records to uniquely identify which entity is associated with each transaction record. Also, with data records being provided by sources from many countries, in many industries, and across many languages the merging challenge is increased. A way to meet this challenge today is to perform processing of the text that is present in the records to determine which records should be merged under an entity. Various techniques of text association, filtering, character grouping, thesaurus lookup, machine learning, natural language processing, search-based comparison, classification, known entity matching, clustering, and the like may be applied to identify mergable records. The complexity and challenge present in merging may require applying each technique in an intelligent way so that highly computation intensive processes such as classification are used appropriately.
One objective of merging is to take a set of input records and match them to entities that are already known to the platform, such as entities already included in an entity database or other database of the platform. When a match cannot be determined automatically with a sufficient confidence level, then information may be presented to an operator or user of the platform to make a final determination of the entity associated with the record(s).
The methods and techniques for identifying mergable records may be programmed into a processing unit and run in a sequence that facilitates rapid and robust merging of records. At least three types of processing may be performed on records for merging assessment: filtering, classification, and clustering. Each processing type will be described now.
Because classification and clustering may be very expensive in terms of compute/processing time, filtering is applied to distinguish candidate records for classification from records that are unlikely to be mergeable under a single entity. Filtering provides various techniques to help identify only the records that classification may have any chance of merging. Filtering for the purposes of merging may be considered a coarse sort of the records, capturing candidates for classification and passing through those records that appear to be far removed from the captured records. Filtering may be performed by a variety of filter-type algorithms. In one example filtering may be performed by search engine software, such as the open source lucene search engine applied. In another example of filtering, sometimes referred to as “kgram filtering”, several small consecutive strings of characters are captured from each of two records and compared. Kgram filtering may be based on techniques of dynamic programming. In an application of kgram filtering, when a sufficient number of the character strings match between the records, the records may be identified as potential candidates for further processing such as classification and clustering. One benefit to kgram filtering is that it offers the filter designer many options, such as allowing overlapping character strings, defining the length of each character string, determining the quantity of matching strings required to mark the record(s) as classification candidates. In this way, an entity name or entity identifying information (that may include entity name, logo, phone number, address, and the like) need not be an exact match, but instead needs enough matching character strings to exceed a kgram filter threshold. In an example, a kgram filter may compare overlapping character strings (kgram filter group) of 10 characters and may require that at least 10 of the character strings must match (kgram filter threshold) for the record to be identified as a potential candidate for classification and clustering. Because records received and processed by the platform may have information within certain fields that may be incorrectly placed there (a personal name in an entity name field) filtering can be used to quickly separate out records that are incorrect.
Another merge technique is called classification. Classification may be performed on any records, although records that have been identified by filtering as candidates for classification may yield faster and more robust classification results. Because records with non-matching entity information may be records of a single entity, classification uses text, language, mathematical, and other analysis techniques to identify a likelihood that two records are from the same entity. Classification includes a variety of techniques including canonical adaptation, specific cleanups, multi field comparison (name, address, phone number, etc), edit distance algorithms, vector generation, machine learning, decision tree, and the like.
In canonical adaptation, entity information in records is adapted to eliminate differences that should not impact classification. Differences such as abbreviations of words (rd. for road, ave for avenue, CA for California, and the like) can be normalized in the records. Punctuation and other characters that may have minor impact on a classification may be removed or marked to be ignored during various classification and clustering techniques. In addition to canonical adaptation targeted cleanups may be applied to further normalize the data. Cleanups may help to resolve deficiencies in the records such as an incorrect country of origin, which is a common deficiency. Cleanups may be based on information about the domain of the records to further enhance entity identification and merging. Cleanups may be based on geographic or regional knowledge, market verticals, industry norms, and the like. In an example, within a market vertical, variations of textile suppliers names may be applied to quickly align the various names to a normalized or canonical entity name; thereby reducing the degree of complexity that further classification techniques will have to deal with. The result may include less complex mathematical computation. Cleanups maybe targeted at specific aspects of the records, such as entity names, city names, street names, phone numbers, and the like. Any number of these cleanups may be applied sequentially or in parallel to data records to improve mergability of the records.
To account for differences in data entry that may result in a very low classification score, classification techniques herein are applied to individual fields (entity name field, address field) as well as to combinations of fields (entity name+address field) so that a record with an entity name in an address field can still be identified as mergable with other records that have the address in the address field.
Data that has been cleaned or adapted as described above and elsewhere herein may be processed through edit distance metric algorithms such as Wagner-Fischer, Levenshtein, Jaro-Winkler, and the like. The result of which may be a complex vector of numbers that represent dimensions of similarity associated with the various classification techniques applied. The vectors of similarity may be based on other classification and text analysis techniques as may be know to those skilled in the art. All such classification and analysis techniques may be applied to the records by the platform and are included herein.
Machine learning and other artificial intelligence techniques may be applied to determine if similarity vectors of pairs of records identify records that can be merged under a common entity. Through the use of training vectors, and decision tree logic, record mergability may be further assessed and a measure of such mergability may be made available to clustering techniques. The result may include an identification of pair-wise matches among all of the classification candidate records.
Training vectors may be derived from transaction data. A set of transactions may be identified as a training set that may be useful in establishing prediction parameters for associating shipments with attributes such as a type of entity, type of supplier, type of product, product feature or attribute, type of material, and the like. A training set may also be useful for facilitating association of a shipment with an entity by enabling development of prediction parameters that may be used therefore. By identifying candidate relationships between shipments and attributes or entities, training sets of transaction records may reduce the computational load required for comprehensively filtering, classifying and clustering. In an example, a record may be presented to a processing facility such as an analysis facility J32. The processing facility may select prediction parameters based on one or more data fields in the record. Certain fields in the transaction record may be compared to a portion of the prediction parameters to predict an entity to associate to a transaction record.
Prediction of an attribute associated with a transaction or customs record may be useful for rolled up, aggregated, or otherwise cumulative transaction data. Because transaction records may be individual shipment records, aggregated transaction records, rolled up or summarized transaction data, and the like, predicting an attribute that may be associated with rolled up transactions may allow the platform to gain significant benefit from otherwise non-specific data. In an example, US customs records may record each shipment from China as an individual customs transaction record but the transactions may not identify the supplier, just the shipper and buyer. However, China may only provide a rolled up transaction that cumulates similar shipments over a period of time, such as a calendar month. The rolled-up transaction data from China may have some data elements that distinguish it partially, such as a product identifier, source region, shipper, supplier, and the like. The US customs transaction data for a calendar month may be used to identify prediction parameters that may be applied to the China transaction data to predict the supplier. When US customs training set data such as shipment quantity, shipper, and the like are applied to the China data, a supplier may be predicted for the rolled up China transaction data.
An objective of clustering is to cluster as many data records that should be merged under a common entity as possible. Clustering may result in all of the variations of one entity being identified as one entity. A technique for clustering that may be applied is referred to herein as p-percent clustering. In p-percent clustering, a pair match threshold is established and any record that matches at least the threshold percent of records in any given cluster will be added to the cluster. In this way, although pair-wise matching identifies all pairs of matched records, clustering allows records that do not all match each other to form a cluster. In an example, if a p-percent threshold is 25% then any record that pair-wise matches at least 25% of cluster members may be added to the cluster. In an embodiment, dynamic p-percent may allow dynamic adjustment of p-percent based on an aspect of the cluster, records, and the like. In an example, p-percent may be set low for a small cluster and may be increased for a large cluster. P-percent clustering ensures that records that have strong matches to some of the members of the cluster can be properly included in the cluster. P-percent offers significant advantages over single dimension (single link) clustering techniques.
Filtering, classification, and clustering are important and facilitate merging of intra data source records (e.g. new transactions for an existing company) as well as external data source records (e.g. US to China customs data records). Also these techniques may be useful in classifying entities into industries or markets.
Industry or vertical classification may be accomplished by using data associated with a shipping record and/or other data sources to determine which industry an entity (buyer, supplier) is associated with. As described above, machine learning techniques such as decision trees can be used to classify individual data records.
The customs transaction data can be mined to automatically build training data for vertical classification. Standardized codes such as the Harmonic Tariff System (HTS) codes embedded in the free text commodity fields can be extracted and used to determine a vertical associated with a record. Along with the HTS code text in the commodity field maybe mined to train a vertical classifier facility. The vertical classifier facility can then be used to predict or determine a vertical classification of customs records. In an example, a commodity field of a record may be “HTS 6209180 Red cotton pants”. —The extracted HTS code 6209180 may be determined to be associated with a garments industry vertical. The extracted label “red cotton pants” may be recognized as apparel in our training data. If “red cotton pants” is not recognized, then it can be added to the apparel training data. Generally only a small fraction of customs data has HTS codes; therefore training a classifier and applying the trained commodity entries to new records may facilitate classification of the remainder of the transaction record. Because the vertical classifier may be a self-learning facility, each new record processed by the classifier can enhance the vertical classifier ability to classify new records. In addition, hand-labeling of records may be used to improve the vertical classifier training data.
In embodiments, a name updater may provide tools to clean the name of a supplier or buyer, such as making commas, periods, capitalization of acronyms, fixing common misspellings, making common abbreviations, and the like consistent. This may be an automated process of cleaning up those names, as well as a manual interface to go through groups of names by glancing at them.
Weights may be applied to rating algorithms, data, and the like in the methods and systems disclosed herein. Weights may be applied in the process of determining ratings so that certain factors that affect a rating may have a greater impact on a rating than other factors. Weights may also be variable and may be based on a combination of factors. Weights may be applied based on a timeliness of data. Timeliness of data may be important to be weighted because, for instance, very new data may not yet be verified or old data may no longer represent a buyer-supplier relationship. Weighting of data may also be important because some data may be of suspect quality independent of age, data may not have a high degree of relevance to a rating, and many other data quality related factors. In this way, weights may be given in the rating process based on timeliness of data, size of transaction, quality of the transacting parties, prior rating of a transacting party or entity, relevance of the data. Weighting factors may be based on human-aided assessment of an entity, financial health of an entity, and the like.
An overall rating 110 of a supplier or buyer may be a combination of sub-ratings such as rating associated with amount of experience, certification dimensions, county context, business metrics, customer loyalty, and the like. An overall rating 110 and any sub-rating may be weighted, normalized, and curve fitted to ensure the rating is providing a consistent reliable measure of a supplier, buyer, and the like. Additionally, the weighting may be customer specified to enable a customer to identify portions of the ratings that are most important.
One sub-rating metric is customer loyalty. Determining a customer loyalty rating for a supplier is computationally intense and algorithmically rich because it measures how well a supplier is at keeping customers, or how well a buyer sticks with a supplier. In some industries, such as in apparel, it is quite common for buyers to change almost half of their suppliers every year. Understanding the factors that determine how this activity impacts customer loyalty is a key benefit of the present invention.
Customer loyalty may be determined by looking at individual buyer-supplier pairs. One technique to determine a customer loyalty rating for a supplier is to determine a customer loyalty rating for each buyer (customer) of that supplier and then combine the individual ratings. Factors that may impact customer loyalty include, a buyer buying pattern, buying frequency, number of purchases, time since first purchase, and the like. Each transaction can be analyzed to determine if the buyer is buying from a second supplier and if the purchase is for an item that was previously purchased from a first supplier. In this situation, customer loyalty of the first supplier is compromised. However, simply measuring transactions may not provide a quality measure of customer loyalty. Factors such as if the first supplier has stopped selling the item that the buyer is now buying from the second supplier are important to include.
Transaction data may often only be available as free text data (UPC and other codes may not be included in the records). Therefore the text processing, normalization, and canonical adaptation techniques described herein may be beneficially applied to determining a customer loyalty rating for a supplier. Certain aspects of buyer-supplier relationships may have greater importance than others so exponential weighting on some dimensions may be useful. In an example, a longer relationship of fewer transactions may be more important than a large number of transactions over a shorter duration. Factors included in a customer loyalty calculation include duration of relationship, count of orders/shipments, the weight of each order/shipment (determines size/value of shipment), and the like. In an example if a buyer buys the same item from two suppliers and consolidates orders to just one of the two suppliers, a customer loyalty rating of the other supplier may be significantly impacted because of the known cutoff in the supplier-buyer relationship.
For merging records, determining an overall rating 110, and other activities and results associated with the platform, determining parent-subsidiary relationships may be important. In addition to parent-subsidiary relationships, other relationships may be important in determining overall rating 110, customer loyalty rating, and the like. A buyer that switches from one subsidiary to another subsidiary under a single parent may have little impact on the parent rating, but may have significant impact on the subsidiary rating.
An aspect of the platform that facilitates determining parent-subsidiary relationships may use various sources of information such as business records from Dunn and Bradstreet, web news feeds, search engine results of business news sites, crawling of supplier web sites, press releases of suppliers, and the like. An acquisition of a subsidiary by a parent may be identified through one or more of these data sources and the parent-subsidiary relationship may be factored into overall rating, customer loyalty rating, merging, and the like. Parent-subsidiary relationships may also be determined based on predetermined heuristics such as same city—similar name, same buyer—similar name, and other heuristic combinations of customs data record elements. Parent-subsidiary relationships can be determined for suppliers and for buyers.
Referring to
Referring to
A name chooser tool, such as the one described above and depicted in
Referring to
Referring to
FIGS. 24 and 25A-B show a breakdown of supplier shipment history, where shipment history may be broken down by piece count, by month, by month to a certain country, and the like.
Referring to
In an embodiment, the user interface 2722 may facilitate a tuple-based search 2748. The tuple-based search 2748 relates to a capability of searching for entities 2708 related to a specific parameter. Such parameter may relate to a product 2750, a material 2752, and/or a technique 2754. In an example, a supplier S1 may like to conduct a search for buyers available in the United States for ‘Aluminum based packaging sheets formed by extrusion.’
In accordance with an embodiment of the present invention, the search results obtained from the above described searches for the entities 2708 may also be ranked. In an embodiment, the ranking may be based on a supplier rating.
In an embodiment, the rating may be based on the context of a party, the business legitimacy of a party, an assessment based on the trading environment of a country, macroeconomic information, industry awards, industry certifications, amount of experience, number of shipments, duration of experience, size of transactions, extent of domestic experience, extent of international experience, caliber of customers, customer loyalty, degree of specialization, specialization in product categories, specialization in manufacturing techniques, specialization in materials, specialization in gender, feedback from customers, feedback from buyers, feedback on product quality, feedback on customer service, feedback on timeliness of delivery, feedback on language skills, feedback on sample making ability, respect for intellectual property, quality management, social responsibility, environmental responsibility, standards of compliance, certifications, certifications with respect to specific vendor standards, risk profile 2758, opportunity profile 2760, and some other types of factors and parameters.
Referring to the above example again, upon searching, a supplier 2732 may obtain a list of buyers 2730, which may be interested in buying ‘Aluminum based packaging sheets formed by extrusion’. In addition, the supplier may like to ascertain the best buyers. For this purpose, the supplier 2732 may also obtain a ranking of the buyers 2730 based on selected parameters such as feedback reports, risk associated with each buyer, geographical location, and some other type of parameter. The rating may be in the form of a value, integer, percentage, and some other forms of ratings. Based on this rating, a ranking may be provided to each of the buyers 2730. This ranking may in turn facilitate the supplier 2732 in making the judgment regarding the appropriate buyer. Risk may be related to counterfeiting, capacity, subcontracting, political factor, geographic factor, weather factor, geology factor, financial risk, probability of non-performance of a contract, probability of termination of a contract, intellectual property, targeted delivery date, transactional customs data for a party other than the supplier and/or buyer, likelihood that a buyer will move to an alternative supplier, non-payment and some other types of risk factors.
An opportunity profile 2760 may be an assessment of the potential for new business opportunities determined from customs transaction data. By analyzing transactions in customs data, buyers and suppliers can identify potential business opportunities such as to establish a new relationship, reduce costs, increase availability, and the like. While companies guard much of their internal information related to costs and profit, the transaction information available in public customs records can provide great insight into ongoing buy and sell activity. In an example of opportunity profile 2760 assessment, a buyer may decide there is an opportunity to push a supplier harder to reduce a price. The buyer may be able to determine that the supplier has made fewer sales (e.g. as evidenced by lower shipment quantities in customs transaction records) over time. One potential reason for this is that a competitor of the supplier is offering a lower price. Therefore the supplier may need to reduce price to remain competitive. 27ikewise the supplier can review the same records and determine that the competitor is selling at a lower price under certain conditions, so the supplier can device a counter pricing strategy accordingly. In another example, a supplier may spot an opportunity to sell additional types of products to an existing buyer by examining the transactions of the buyer. The supplier may determine that the buyer is purchasing a type of product from a competitor that the supplier also offers but is not currently selling to the buyer. The supplier could provide the buyer with the opportunity to potentially improve the buyer costs by ordering the product from the supplier rather than the competitor. Factors such as combined volume pricing, reduced accounting overhead, lower shipping costs and the like may be key benefits that the supplier can use to entice the buyer.
Likewise the platform or an operator of the platform may use the customs transactional data to identify and suggest opportunities to buyers and/or suppliers. The transactional data may be analyzed for factors that indicate the potential for an opportunity and the opportunity may be prepared as an offer to one or more of buyers, suppliers, and the like. Opportunities may include availability of pricing leverage for a buyer with respect to a supplier; consolidation of orders with a supplier, pricing leverage for a supplier with respect to a buyer, increasing a share of a buyer's total spending for a supplier, and the like.
A risk profile 2758 may be determined based on analysis of customs transaction data. A risk profile for a supplier or a buyer may be based on customs transaction data for the supplier, the buyer, or a third party. A risk profile that may be determined from customs transactional data may include risk related to counterfeiting, capacity overload, subcontracting, political factors, geographic factors, weather, geology, finances, probability of non-performance to a contract, probability of termination of a contract, intellectual property, achieving a targeted delivery date, non-payment, selecting an alternate supplier, order cancellation, order push-out, and the like. A risk profile that may be derived from customs transaction data may be a basis for determining terms and conditions of insurance, and the like.
The above description disclosed that the search interface may be utilized for searching entities 2708 based on the aggregated transactional data. In an embodiment, the suppliers 2732 may also be searched based on the region of interest (geography) 2738, industry specialization 2740, customers (entity types) 2742, and the interest displayed in forming relationship (likelihood of interest) 2744. This may be explained in detail in conjunction with
Referring to
In embodiments, the search facility 2720 may be adapted to be used by a buyer for searching a supplier. Alternately, the search facility may also be adapted to be used by a supplier for searching a buyer.
In an example, a buyer 2730 may like to search for suppliers (in US) of ‘automotive machine parts’ that may be willing to do business with a small offshore firm outside United States. Therefore, in the above scenario the likelihood of interest that the suppliers may display may be based on the location and size of the firm.
In embodiments, methods and systems disclosed herein may include an interface by which buyers may search for suppliers, as disclosed above. The search interface may allow buyers to query a database of supplier information organized in a hierarchy according to product categories, in order to find suppliers who provide products in a selected category. A buyer may then select particular suppliers and obtain an online profile or report, as described throughout this disclosure, as to attributes of a particular supplier. In embodiments, the search interface allows the buyer to search by product category, material used to make the product, or technique used to make the product, among other attributes. Filtering tools may be provided in the interface to allow the buyer to sort data by product type, material, technique, caliber of customer, or other attributes, to expand or group data, to drill down into particular categories or sub-categories, and the like.
In embodiments, the database of supplier information includes an ontology of product categories, which may include a tree of categories and sub-categories of all types of products found in various data sources, such as customs records databases.
In embodiments, filters may be enabled, allowing a buyer to search along dimensions of the data. In an example, if a buyer wishes to search for suppliers who work with a particular material, a filtering algorithm may take the union of all materials used by suppliers and present those materials as filters by which a set of suppliers may be selected by buyers for further analysis. The filters may be presented in a graph or tree structure, so that a user may check a box to expand or contract a particular portion of the tree, thereby allowing filtering by sub-category down to the leaf node in a tree. In embodiments, data are represented in tuples and results for a particular filter are ordered, such as by overall rating of the supplier. Results for a particular filter may also be ordered by other features, such as most specialized and the like.
Filters may include construction techniques, dyeing, washing and embellishing techniques, gender of the product, company type, country of supplier, and the like. When data are represented in tuples, all products a supplier has made may be represented by material, sub-material, and technique (e.g., cotton—poplin—knitted sweater). In an example, when a search is conducted for a cotton poplin sweater, the suppliers who have made cotton poplin sweaters can be retrieved (not the union of ones who have made cotton or poplin sweaters in this example). The tuple concept applies to children of each concept in a hierarchy, so if the user selects cotton, the user will receive results for cotton and all children of cotton in the materials hierarchy.
In embodiments, the search interface may include a non-tuple-based search mode in which suppliers would be suggested as possible matches for the user's query that would be the union of the search terms. In an example, if a supplier has worked with silk, and has produced pants, the system predicts that this supplier could make silk pants.
In embodiments, a user interface may include paid or sponsored links in addition to search results derived from the rating platform described herein.
In embodiments, methods and systems disclosed herein may include private and public versions of reports, where a searcher can get to a public profile by an Internet site but requires some additional relationship (possibly involving payment) in order to drill down to receive more information, such as a complete profile of a supplier.
In embodiments, various icons, filters, sliders or other techniques may be provided in a user interface to allow a user to explore information about a supplier. In embodiments, a user can click for “details,” thereby pulling up a ranking a supplier has for a given dimension, with information about the data source and a reminder of the purpose of that dimension. In embodiments, icons may show if a score is high, medium or low, thereby bucketing suppliers into general categories. In embodiments, filters or sliders may allow users to refine results, such as to show suppliers only if the product in question represents at least a minimum percentage of that supplier's product mix.
In embodiments, an interface may be provided for rating a supplier, such as on dimensions including an overall rating, product quality, customer service, timeliness, English language capability, sample-making ability, respect for intellectual property, and the like. Buyer ratings may be averaged or otherwise normalized and reported as part of a supplier's overall rating 110. In embodiments, transactional data may be used to ensure that a transaction occurred (to keep ratings unpolluted). If a buyer rating is good, this can give a significant boost to an overall rating 110.
In embodiments, buyers could specify which dimensions are most important to them, and the overall rating 110 could be customized and weighted according to the buyer's preferences.
In embodiments, suppliers may be suggested to buyers based on the types of qualities the buyer seems to appreciate, and the types of products the buyer has produced in the past.
The capability to identify and classify various buyers and suppliers as ‘friends’ or the like may also be facilitated by using public transaction records 2804, as shown in
Referring to
In embodiments, a buyer may identify like (similar) buyers. Similarly, a supplier may identify similar suppliers.
In other embodiments, a supplier may identify buyers like those of the supplier. A buyer may also identify suppliers like those of the buyer.
In embodiments, a buyer may identify suppliers of a specified type. Further a supplier may identify buyers of a specified type.
In embodiments, a buyer may identify suppliers most likely to prefer a particular buyer. Similarly, a supplier may identify buyers that would prefer a specific supplier.
In embodiments, the public records of transactions 2804 may also be used for classification of buyers 2838. This has been explained in conjunction with
In embodiments, an interface may include a capability for buyers to network, chat or otherwise interact with each other with respect to suppliers. Such a network may include a capability of identifying other buyers as “friends” or the like, thereby allowing sharing of information only among trusted parties. In such a case, information about suppliers for particular buyers might be automatically populated, simplifying the sharing of information about experiences with particular suppliers used by a network of buyers.
In embodiments, analyses could be used to assess and/or identify credit worthiness of suppliers or buyers.
In embodiments, ratings could be embedded into other media such as other websites, emails, print media, and the like. Such embeddings could be a result of calls to an Application Programming Interface (API), or other methods.
In embodiments, ratings may be grouped into buckets, such as “excellent,” “good,” “fair,” “poor,” and “not trade worthy.” Various methods may be used to group suppliers into such buckets. In an example, “excellent” ratings may be given to suppliers who have business legitimacy and are in the top quartile in both loyalty and experience, “good” ratings may be given to suppliers who have business legitimacy and are in the top half in both loyalty and experience, “fair” ratings may be given to suppliers who have business legitimacy and are in the top half in either loyalty or experience, “poor” ratings may be given to other suppliers who have business legitimacy, and “not trade worthy” ratings may be given to suppliers who do not have indicia of business legitimacy.
In embodiments, methods and systems disclosed herein may assist suppliers in generating leads among buyers for opportunities to supply products. Information about how to improve ratings may be used to assist suppliers in generating high quality leads.
The methods and systems of the platform may facilitate identifying a supplier of an item type that is disclosed in a transaction record even if the supplier is not a party to the transaction. Identifying a supplier of an item, or a type of item may benefit buyers, suppliers, and the like by identifying potential new relationships between buyers and suppliers. A buyer may use the resulting supplier identification and other information in transaction records, such as declared customs value, to compare a current supplier cost with a different supplier cost for the same item. A buyer or a potential buyer may use at least transaction cost and delivery information to identify suppliers from which the buyer may request a quote for supplying the item. Suppliers may identify buyers of products that the supplier also provides. This may lead to efficient marketing and sales activity for the supplier because the supplier would know that buyer has a significant interest in the item being purchased. By examining other information in the transaction, such as buyer behavior, transaction history and the like, the supplier may identify an offer profile of the buyer and present a very well targeted offer to the buyer.
In accordance with an embodiment of the present invention, the information from the aggregated transactions may also be utilized for supplier assessment. Supplier assessment may involve determining if a specific buyer has ceased business operations with a supplier. Such a determination may be based on a cycle time between shipments which may be based on historical shipment data derived from transaction records. A calculation of cycle time of shipments for an item from a supplier to a buyer may indicate an approximate date of a next shipment. If a transaction record reflecting the next shipment does not show up in the transaction records within some period of time beyond the indicated next shipment date, the methods and systems may indicate that the buyer may have stopped business operations with the supplier. The nature of the stoppage may be further determined if transaction records indicate that the buyer has begun receiving shipments of the item from a different supplier. Cycle time calculations may also be used to evaluate a supplier's delivery performance. Significant increases in cycle time may indicate delay of shipment by the supplier. An assessment of supplier-buyer transaction status may also include factoring in buyer inventory. Buyer inventory may be factored in as a prediction or estimate of inventory.
In an embodiment of the present invention, methods and systems may be provided for rating an entity based on rolled up customs data. Rolled up customs data may include aggregated, cumulative, summary, or similar methods of combining a series of transactions into roll-up data. Rolled-up data may include a total of shipments over a period of time for a buyer-seller-product association. Rolled-up data may include total shipments over a period of time for product-shipper association. Any and all types of consolidation of transaction data that may be based on a time interval, a frequency, a region, an industry, a product, a supplier, a buyer, a shipper, a source region, an exchange rate, and the like are herein included. In an example, rolled-up transaction data may include a supplier's total output in each product category over a calendar month. In another example a country may report a total export of a product during a week. In both cases critical transaction information that may be missing may be estimated or predicted in order to develop otherwise useful information from the rolled-up information. The computer implemented facility 2902 may collect and store the rolled-up public records of transactions 2904 and aggregate them to form aggregated transactions.
In an embodiment, the transaction records may relate to the shipment transactions.
Further, the aggregated transaction records may be associated with a particular supplier. This associated information may be analyzed to determine and convey the rating for the supplier.
It may be appreciated that this procedure may be conducted periodically (In an example, every three months). In another embodiment, the change in rating of a specific supplier may be presented as an alert.
In an example, the embodiments described above may be utilized by a company dealing in an improved form of pesticide that may wish to determine the ratings of a specific supplier of raw materials situated in a different country. Therefore, the aggregated and associated shipment transaction information (regarding the shipment time, schedule, price and delivery) for the supplier may be used to determine its rating among a plurality of similar suppliers. Subsequently, this rating may be instrumental in helping the above company make supply related business decisions.
The public records of transactions may also be utilized for predicting minimum order requirements for a factory. Referring to
In an embodiment, the entity may be a factory.
In another embodiment, the entity may be a supplier.
In yet another embodiment, the entity may also be a subsidiary of a supplier.
In an example, information regarding the public transaction records 3004 such as transaction receipts for a candle supplier selling a batch of factory-made candlesticks from a candle manufacturer may be aggregated and associated by the computer implemented facility 3002. The analysis facility 3020 may perform detailed analysis of this information to generate various types of results. In an embodiment, the analysis facility 3020 may predict the minimum order requirement for an entity, based on the analysis of the transactions. As described in the above example, the analysis facility 3020 may predict the number of batches that the candle manufacturer may need to sell in order to cross the minimum profit mark. In another example, the analysis facility 3020 may predict the minimum number of candle-stick batches that may need to be supplied to a third party in order to fulfill the required terms laid down in a mutual contract. In yet another scenario, the analysis facility 3020 may also facilitate predicting the minimum order requirements that a subsidiary of a supplier may need to supply among the batch of suppliers.
In embodiments, methods and systems are disclosed herein for using disparate data sources, including transactional records, such as from customs transactions, as a basis for rating suppliers of products. In embodiments, transactional data from actual transactions are used to generate experience ratings, specialization ratings, customer loyalty ratings, or other ratings.
In embodiments, a rating system is provided in which buyers rate suppliers of products, wherein transactional data, such as from customs records, are used to verify the legitimacy of the feedback, such as to verify that a rated transaction actually occurred.
In embodiments, methods and systems allow buyers to search for suppliers, including with filters based on product category, material or techniques offered or used by the suppliers, and to retrieve ratings information about the suppliers, including ratings derived from transactional data (such as customs data) or ratings derived from other buyers.
In embodiments, a platform for enabling searches for suppliers and ratings of suppliers may include various tools, such as tools for merging records, merging supplier names, and the like.
In embodiments, methods and systems disclosed herein may include a quote tool by which buyers may identify suppliers and then generate a request for a quote from selected suppliers.
In embodiments, algorithms may be used for determining a pricing leverage metric, such as based on transactional data, such as customs records. In an example, a supplier's pricing leverage may depend on the percentage of a supplier's shipments that are going to a single buyer, the proximity to a recent switch in supplier by one or more buyers, a supplier's overall score, a supplier's customer loyalty score, a supplier's experience in an area, and global factors, such as overall demand for a product offered by a supplier. Thus, an interface may allow buyers to assess pricing leverage based on calculations using one or more of these factors, normalized or weighted to provide an overall estimate or score as to pricing leverage of a supplier.
In embodiments, access to the database may be restricted by the capacity of a supplier, and the ability of a supplier to ship small quantities. For instance, large buyers may need access to the entire database, while smaller buyers may need special access to suppliers that specialize in smaller orders.
A user interface may include various alerts, such as an alert for when a new supplier satisfies a search criterion of a buyer.
Methods and systems disclosed herein may include methods for syndicating data, such as delivering overall scores, category ratings (e.g., “excellent,” “good,” “fair,” or “poor”) or the like, to third parties, such as for presentation in connection with other business data, such as data presented to securities analysts, data presented to buyers for other purposes, and the like. Users will be able to provide subjective ratings of suppliers on such third party presentations using an API.
In embodiments, methods and systems disclosed herein may include collaborative filtering techniques, such as to allow buyers to see information relevant to other buyers who share characteristics with the buyer (such as conducting similar searches, using similar suppliers, or having similar transaction records). Collaborative filtering may also allow suppliers to access information relevant to other suppliers with similar characteristics, such as ratings for suppliers to supply the same types of products, to the same types of buyers, and the like.
In embodiments, a buyer scorecard 200 may be provided that shows summary data for a supply chain of various suppliers, such as to indicate how the buyer's suppliers collectively compare to suppliers of other buyers, such as competitors of the buyer.
In embodiments, a supplier comparison tool may be used to compare suppliers on various attributes.
In embodiments, buyers may be rated on behalf of suppliers, such as based on loyalty to suppliers.
In embodiments, ratings of suppliers or buyers as described throughout this disclosure may be used for third parties, such as, in embodiments, financial analysts. In an example, an analyst could evaluate the quality of a company's supply chain based on collective supplier ratings. Similarly, an insurance company could use data about suppliers of a buyer to assess supply chain risk, such as for analyzing risk associated with insurance associated with activities of suppliers.
In embodiments, buyers may supply data to the platform described herein in order to assist with developing ratings, but that data may be maintained as proprietary to the buyer, such as to keep ratings generated based on that data private to the buyer.
In embodiments, information about suppliers may be syndicated to desk software tools, such as tools used by purchasing managers and buying staff within buyer organizations. Thus, reports or ratings may be fed so that they appear within the interface of one or more other desktop or web-based tools used by such users.
In embodiments, methods and systems disclosed herein may include filtering tools for sorting data retrieved from customs records according to an industry hierarchy, such as a hierarchy of products, materials and techniques.
In embodiments, a search interface may allow for a search based on supplier capability, such as based on information retrieved from transactional data, such as customs records.
In embodiments, a data analytics platform may be provided for analyzing supplier capabilities, such as based at least in part on transactional data about supplier activities, such as transactional data from customs records.
In embodiments, a rating system may be based on a combination of customs data and other data, such as data based on an internal database of transactions made by an agent on behalf of buyers transacting with suppliers.
In embodiments, a platform may include a transactional facility, such as for allowing buyers to transact with suppliers that have been identified by the search and ratings facilities described herein. Such transactional facility may include modules related to ordering, pricing, payment, fulfillment, and the like.
Referring to
In an example, the aggregated transactions information may reveal a .list of twenty entities doing business in an uptown market. A searcher may utilize the methods and systems disclosed herein to determine a list of seven entities that may be sub-entities for a specific supplier S1. In accordance with the embodiments of the present invention, these seven entities may be rated based on the transactional data. The seven entities (say 2 factories, 3 subsidiaries, and 2 sales divisions) may be rated based on the timeliness of the delivery, feedback from the buyers, and so on.
In an embodiment, the analysis facility 3124 may determine the sub-entities 3140 for a supplier from the group of entities. The determination of sub-entities may be based on the analysis of the public transaction records 3104. In an embodiment, the public transaction records 3104 may be customs transaction records.
In another embodiment, the sub-entities 3140 of the supplier may be rated based on the analysis of the aggregated transactions and other information and parameters as explained throughout the disclosure.
The elements depicted in flow charts and block diagrams throughout the figures imply logical boundaries between the elements. However, according to software or hardware engineering practices, the depicted elements and the functions thereof may be implemented as parts of a monolithic software structure, as standalone software modules, or as modules that employ external routines, code, services, and so forth, or any combination of these, and all such implementations are within the scope of the present disclosure. Thus, while the foregoing drawings and description set forth functional aspects of the disclosed systems, no particular arrangement of software for implementing these functional aspects should be inferred from these descriptions unless explicitly stated or otherwise clear from the context.
Similarly, it will be appreciated that the various steps identified and described above may be varied, and that the order of steps may be adapted to particular applications of the techniques disclosed herein. All such variations and modifications are intended to fall within the scope of this disclosure. As such, the depiction and/or description of an order for various steps should not be understood to require a particular order of execution for those steps, unless required by a particular application, or explicitly stated or otherwise clear from the context.
The methods or processes described above, and steps thereof, may be realized in hardware, software, or any combination of these suitable for a particular application. The hardware may include a general-purpose computer and/or dedicated computing device. The processes may be realized in one or more microprocessors, microcontrollers, embedded microcontrollers, programmable digital signal processors or other programmable device, along with internal and/or external memory. The processes may also, or instead, be embodied in an application specific integrated circuit, a programmable gate array, programmable array logic, or any other device or combination of devices that may be configured to process electronic signals. It will further be appreciated that one or more of the processes may be realized as computer executable code created using a structured programming language such as C, an object oriented programming language such as C++, or any other high-level or low-level programming language (including assembly languages, hardware description languages, and database programming languages and technologies) that may be stored, compiled or interpreted to run on one of the above devices, as well as heterogeneous combinations of processors, processor architectures, or combinations of different hardware and software.
Thus, in one aspect, each method described above and combinations thereof may be embodied in computer executable code that, when executing on one or more computing devices, performs the steps thereof. In another aspect, the methods may be embodied in systems that perform the steps thereof, and may be distributed across devices in a number of ways, or all of the functionality may be integrated into a dedicated, standalone device or other hardware. In another aspect, means for performing the steps associated with the processes described above may include any of the hardware and/or software described above. All such permutations and combinations are intended to fall within the scope of the present disclosure.
While the invention has been disclosed in connection with the preferred embodiments shown and described in detail, various modifications and improvements thereon will become readily apparent to those skilled in the art. Accordingly, the spirit and scope of the present invention is not to be limited by the foregoing examples, but is to be understood in the broadest sense allowable by law. All documents referenced herein are hereby incorporated by reference.
Claims
1. A method, comprising:
- using a computer implemented facility to collect and store a plurality of customs transactional records;
- inferring an entity identifier for each of a portion of the plurality of customs transactional records;
- processing with the computer implemented facility data from each of the portion of the plurality of customs transactional records to determine an entity name that is representative of a plurality of different inferred entity identifiers based on similarity of data among the plurality of customs transactional records;
- aggregating the customs transactional records of the plurality of different entity identifiers for at least one representative entity name; and
- facilitating access to at least one of a portion of the aggregated customs transactional records and information that is derived from the portion of the aggregated customs transactional records.
2. The method of claim 1, wherein facilitating access is via the entity name.
3. The method of claim 1, wherein at least one entity identifier is a variation of the entity name.
4. The method of claim 1, wherein facilitating access comprises presenting the information that is derived from the portion of aggregated customs transactional records in an electronic user interface in response to a user entering a query through the interface comprising at least one of an entity name, an entity identifier, and data that is a variation of the entity name.
5. The method of claim 1, wherein the information that is derived is representative of shipping performance of the determined entity name.
6. The method of claim 1, wherein the information that is derived is a rating of the determined entity.
7. The method of claim 1, wherein the entity name is for one of a buyer, supplier, and shipper.
8. The method of claim 1, wherein the determined entity name is one of the different entity identifiers.
9. The method of claim 1, further comprising integrating at least one of the aggregated customs transactional records with data from at least one other data source.
10. The method of claim 9, wherein facilitating access comprises facilitating access to at least one of a portion of the aggregated customs transactional records integrated with data from at least one other data source and information that is derived from the portion of the aggregated customs transactional records integrated with data from at least one other data source.
11. A system, comprising:
- a computer implemented facility adapted to collect and store a plurality of customs transactional records, wherein the plurality of customs transactional records comprises a plurality of data fields from which an entity identifier for an entity associated with each transactional record is inferred by the computer implemented facility;
- a module of the computer implemented facility for executing an algorithm that can determine an entity name that is representative of a plurality of different inferred entity identifiers based on similarity of data among the plurality of stored customs transactional records;
- an electronic user interface for receiving a user query that comprises entity information that comprises at least one of an entity name, an entity identifier, and data that is a variation of the entity name;
- an entity search facility for processing the user query to facilitate access to at least one of a portion of the customs transactional records and information that is derived from the customs transactional records based on a similarity of the entity information in the query to at least one of the inferred entity identifiers and the determined entity name.
12. The method of claim 11, wherein the computer implemented facility is adapted to aggregate the plurality of customs transactional records with entity identifiers that are represented by the determined entity name.
13. The method of claim 12, wherein the entity search facility facilitates access to the aggregated customs records.
14. The method of claim 11, wherein to facilitate access by the entity search facility comprises presenting user query search results in the electronic user interface.
15. The method of claim 14, wherein the user query search results comprise at least one of a portion of the customs transactional records and information that is derived from the customs transactional records based on a similarity of the entity information in the query to at least one of the inferred entity identifiers and the determined entity name
16. The method of claim 11, further comprising a data integration facility for integrating at least one of the aggregated customs transactional records with data from at least one other data source.
17. The method of claim 16, wherein the entity search facility facilitates access to at least one of a portion of the aggregated customs transactional records integrated with data from at least one other data source and information that is derived from the portion of the aggregated customs transactional records integrated with data from at least one other data source.
Type: Application
Filed: Dec 5, 2013
Publication Date: Apr 3, 2014
Applicant: Panjiva, Inc. (New York, NY)
Inventors: James Ryan Psota (Cambridge, MA), Joshua Green (New York, NY)
Application Number: 14/097,474
International Classification: G06Q 30/06 (20060101);
