Abstract: Systems and methods for customizing business applications based upon user intent scores is described. A machine learning model trained to specifically predict when a user is likely to engage in a specific activity while interacting with the business application may be trained using data regarding prior interactions between a business application and a plurality of users. The machine learning model may thereafter provide a probability score for a particular user, the probability score indicating the likelihood that the user will engage in the specific activity for which the model has been trained to predict. The probability may be combined with a business value factor to produce a user intent score indicating the relative value of the user engaging in the specific activity.

Abstract: Computer-implemented methods and systems are provided for generating a distributed representation of electronic transaction data. Consistent with disclosed embodiments, generation may include receiving electronic transaction data including first and second entity identifiers. Generation may also include generating an output distributed representation by iteratively updating a distributed representation using the electronic transaction data. The distributed representation may include rows corresponding to first entity identifiers and rows corresponding to second entity identifiers. An iterative update may include generating a training sample and an embedding vector using the components and the distributed representation; determining, by a neural network, a predicted category from the embedding vector; and updating the distributed representation using the predicted category and the training sample.

Abstract: Computer-implemented methods and systems are provided for generating a distributed representation of electronic transaction data. Generation may include receiving electronic transaction data including first and second entity identifiers. Generation may also include generating an output distributed representation by iteratively updating a distributed representation using the electronic transaction data. The distributed representation may include rows corresponding to first entity identifiers and rows corresponding to second entity identifiers. An iterative update may include generating a training sample and an embedding vector using the components and the distributed representation; determining, by a neural network, a predicted category from the embedding vector; and updating the distributed representation using the predicted category and the training sample. The embodiments may also include outputting the output distributed representation to determine authorization of electronic transactions.

Abstract: Computer-implemented methods and systems are provided for generating a distributed representation of electronic transaction data. Generation may include receiving electronic transaction data including first and second entity identifiers. Generation may also include generating an output distributed representation by iteratively updating a distributed representation using the electronic transaction data. The distributed representation may include rows corresponding to first entity identifiers and rows corresponding to second entity identifiers. An iterative update may include generating a training sample and an embedding vector using the components and the distributed representation; determining, by a neural network, a predicted category from the embedding vector; and updating the distributed representation using the predicted category and the training sample. The embodiments may also include outputting the output distributed representation to determine authorization of electronic transactions.

Abstract: Computer-implemented methods and systems are provided for generating a distributed representation of electronic transaction data. Consistent with disclosed embodiments, generation may include receiving electronic transaction data including first and second entity identifiers. Generation may also include generating an output distributed representation by iteratively updating a distributed representation using the electronic transaction data. The distributed representation may include rows corresponding to first entity identifiers and rows corresponding to second entity identifiers. An iterative update may include generating a training sample and an embedding vector using the components and the distributed representation; determining, by a neural network, a predicted category from the embedding vector, and updating the distributed representation using the predicted category and the training sample.

Abstract: Computer-implemented methods and systems are provided for generating a distributed representation of electronic transaction data. Consistent with disclosed embodiments, generation may include receiving electronic transaction data including first and second entity identifiers. Generation may also include generating an output distributed representation by iteratively updating a distributed representation using the electronic transaction data. The distributed representation may include rows corresponding to first entity identifiers and rows corresponding to second entity identifiers. An iterative update may include generating a training sample and an embedding vector using the components and the distributed representation; determining, by a neural network, a predicted category from the embedding vector; and updating the distributed representation using the predicted category and the training sample.

Abstract: An event type generator may provide a training set for classifying documents with respect to an event type. The event type generator may include a request handler to receive the event type and at least one example document, a text analyzer to extract first entities from the at least one example document, and a result manager to execute a first search against an indexed corpus of documents, to obtain first search results, and further to receive at least one selected document from the first search results. The request handler may extract second entities from the at least one selected document, and execute a second search against the indexed corpus of documents, to obtain second search results. The event type generator may thus provide the at least one example document, the first search results, and the second search results as the training set.

Abstract: A document is received that has a plurality of lines with text. This document includes text associated with at least one topic of interest and text not associated with the at least one topic of interest. Thereafter, it is determined, for each line in the document, a length of the line and a number of off-topic indicators with the off-topic indicators characterizing portions of the document as likely being not being associated with the at least one topic of interest. Thereafter, a density for each line can be determined based on the determined line length and the determined number of off-topic indicators. The determined densities for each line are used to identify portions of the documents likely associated with the at least one topic of interest so that data characterizing the identified portions of the document can be provided. Related apparatus, systems, techniques and articles are also described.

Abstract: A system may include a record generator to receive a plurality of documents associated with a plurality of suppliers and provide supplier-specific data records based thereon. The record generator may include an event classifier configured to execute a supplier-independent, event-based classification of each document, to thereby obtain event-classified documents. The record generator may include a supplier query generator configured to query the plurality of documents to obtain potential supplier matches from the plurality of suppliers, and a supplier match analyzer configured to analyze each potential supplier match of the potential supplier matches, to thereby obtain supplier matches. The record generator may include a supplier relevance analyzer configured to relate, for each event-classified document, any supplier identified therein to at least one event of the event-classified document, to thereby obtain supplier-event relationships.

Abstract: A system and method of record matching using regular expressions and finite state representations. In this manner, the time (or computational effort) involved in record matching is reduced.

Abstract: A system may include a record generator to receive a plurality of documents associated with a plurality of suppliers and provide supplier-specific data records based thereon. The record generator may include an event classifier configured to execute a supplier-independent, event-based classification of each document, to thereby obtain event-classified documents. The record generator may include a supplier query generator configured to query the plurality of documents to obtain potential supplier matches from the plurality of suppliers, and a supplier match analyzer configured to analyze each potential supplier match of the potential supplier matches, to thereby obtain supplier matches. The record generator may include a supplier relevance analyzer configured to relate, for each event-classified document, any supplier identified therein to at least one event of the event-classified document, to thereby obtain supplier-event relationships.

Abstract: An event type generator may provide a training set for classifying documents with respect to an event type. The event type generator may include a request handler to receive the event type and at least one example document, a text analyzer to extract first entities from the at least one example document, and a result manager to execute a first search against an indexed corpus of documents, to obtain first search results, and further to receive at least one selected document from the first search results. The request handler may extract second entities from the at least one selected document, and execute a second search against the indexed corpus of documents, to obtain second search results. The event type generator may thus provide the at least one example document, the first search results, and the second search results as the training set.

Abstract: A system and method of record matching using regular expressions and finite state representations. In this manner, the time (or computational effort) involved in record matching is reduced.

Abstract: A company is associated, in an enterprise resource planning system, with a plurality of business entities that each have at least one structured record used by the enterprise resource planning system to characterize the business entity. Thereafter, documents are obtained from a plurality of information sources that characterize events associated with each business entity. It is then determined, using pre-defined business rules, which of the events are pertinent to the company so that enhancement records can be generated for the events determined to be pertinent to the company. These enhancement records characterize the corresponding event and are linked to the structured record for the corresponding business entity. Related apparatus, systems, techniques and articles are also described.

Abstract: Data is received that comprises an entity name. Thereafter, it is determined (i) whether there are any punctuation variations for the entity name, (ii) whether there is at least one character to drop from the entity name, and (iii) whether there are alternative equivalents of at least a portion of the entity name. After such determinations have been made, a plurality of variants for the entity name is generated based on a combination of each determined punctuation variation, determined at least one character to drop, and determined alternative equivalent. Related apparatus, systems, techniques and articles are also described.

Abstract: A document is received that has a plurality of lines with text. This document includes text associated with at least one topic of interest and text not associated with the at least one topic of interest. Thereafter, it is determined, for each line in the document, a length of the line and a number of off-topic indicators with the off-topic indicators characterizing portions of the document as likely being not being associated with the at least one topic of interest. Thereafter, a density for each line can be determined based on the determined line length and the determined number of off-topic indicators. The determined densities for each line are used to identify portions of the documents likely associated with the at least one topic of interest so that data characterizing the identified portions of the document can be provided. Related apparatus, systems, techniques and articles are also described.

Abstract: A computer readable storage medium includes executable instructions to convert an entity to a standard form including normalized attributes, a tag reference and a feature. The entity is expanded with corresponding variants. The standard form and corresponding variants are combined to form an annotated entity in a first processing step. The entity is assigned to a group in a second processing step that accesses the annotated entity. The entity is processed in a single pass comprising the first processing step and the second processing step.