Abstract: The present disclosure relates to computer-implemented methods, software, and systems for obtaining incoming alerts associated with tickets generated for health status issues identified by applications for central components of a cloud platform environment. The obtained incoming alerts are aggregated to determine a root cause ticket. A set of incoming alerts is associated with a first central component running at a first data center from one or more data centers of the cloud platform environment. The set of incoming alerts are generated based on a set of tickets logged at a first ticketing system of the first data center by one or more applications running on the first data center. A root cause ticket is generated to combine the set of tickets corresponding to the set of incoming alerts, and the root cause ticket identifies the first central component as associated with a first health status issue.

Abstract: A computer-implemented method, system and computer program product for recommending a template or a dataset. A descriptor (input descriptor) of the input, corresponding to a dataset or a dashboard template, is extracted or fetched. The input descriptor may be the descriptor of the dataset which includes topics and concept combinations. Alternatively, the input descriptor may be the descriptor of the dashboard template which includes topics, concept combinations and column-to-visualization mappings. Scores are then generated for the candidate targets, corresponding to dashboard templates or datasets, based on one or more of the following: matching topics in the input descriptor with topics in the candidate target descriptors, matching concept combinations in the input descriptor with concept combinations in the candidate target descriptors, and determining compatibility of the datasets with the dashboard templates.

Abstract: Systems as described herein may analyze big data to generate predictions for Software-as-a-Service (SaaS) offerings and/or other subscription-based software. A prediction system may receive small business transaction data comprising payment information between a plurality of merchants and a software vendor. The prediction system may receive enterprise merchant insights information. The prediction system may identify, based on the small business transaction data and the enterprise merchant insights information, recurring transactions using a first machine learning model. The prediction system may generate an average contract price and a confidence score, using a second machine learning model. The prediction system may generate a recommendation comprising a renewal rate. Based on a determination that a service provided by the software vendor is coming due for renewal, the prediction system may send, to a particular merchant, an alert associated with the renewal and the recommendation.

Abstract: The present disclosure provides for systems and methods for processing data. In some aspects, a data processing system may comprise at least one computing device, wherein the computing device may be configured to facilitate the performance of at least one process for extracting one or more domain-specific topics or themes from at least one data set. In some implementations, one or more data sets may be received by the data processing system such that the data processing system may identify one or more n-grams within each data item. In some embodiments, one or more of the n-grams may identified as being potentially associated with a correlating topic or theme such that data items comprising the same or similar n-grans may be aggregated into the same topic or theme, thereby assisting a user with quickly locating desirable data items within the data set.

Abstract: The method provides for receiving a plurality of documents including mentions of a target entity from a search query about the entity. The mentions of the target entity are identified in respective documents of the plurality of documents. Content surrounding the one or more mentions of the target entity are extracted with the mentions within the respective documents and form section. A respective document includes a plurality of sections. Metrics of relevance and irrelevance to the target entity are determined within the plurality of sections of the respective documents. A density score is generated for the plurality of sections of the respective documents. A relevancy score is assigned to respective documents of the plurality of documents, based on the density scores of the sections of the respective documents. The documents are ranked based on the relevancy score and presented in an order based on the ranking.

Abstract: Search may be provided using a database storing a plurality of documents comprising a first set of documents and a second set of documents, a set of vetting values and a computer readable medium. In such a system, for each document in the second set of documents, the first set of documents comprises a document for which that document from the second set of documents is identified as a subsequent related document. Additionally, the set of vetting values may comprise, for each document from the second set of documents, a vetting value for the document from the first set of documents for which that document from the second set of documents is identified as the subsequent related document. Additionally, the medium may store instructions to respond to a query by determining, based on the set of vetting values, a search result set comprising documents from the first set of documents.

Abstract: A user can submit a visual query that includes one or more images. Various processing techniques such as optical character recognition (OCR) techniques can be used to recognize text (e.g. in the image, surrounding image(s), etc.) and/or various object detection techniques (e.g., machine-learned object detection models, etc.) may be used to detect objects (e.g., products, landmarks, animals, humans, etc.) within or related to the visual query. Content related to the detected text or object(s) can be identified and potentially provided to a user as search results or a proactive content feed. As such, aspects of the present disclosure enable the visual search system to more intelligently process a visual query to provide improved search results and content feeds, including those search results which are more personalized and/or consider contextual signals to account for implicit characteristics of the visual query and/or user's search intent.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for adaptive educational activities are disclosed. In one aspect, a computer-implemented method includes receiving, by a computing device that is configured to grant a given user access to a first group of activity modules based on a classification of the given user, user data and classification data. The method includes granting the user access to a first activity module and a second activity module, and preventing the user from accessing a third activity module. The method includes receiving interaction data based on the user interacting with the first activity module, and determining that the interaction data indicates an improper classification of the user. Based on the interaction data, the method includes maintaining the classification of the user, granting the user access to a fourth activity module, and preventing the user from accessing the second activity module.

Abstract: Manually framing questions for research and evaluation has the disadvantage that quality and effectiveness of questions of the questions depend on knowledge and expertise of subject matter expert who is framing the questions. The disclosure herein generally relates to data processing, and, more particularly, to a method and system for generating protocol for data extraction from one or more users. The system identifies sub-objectives corresponding to an identified research objective, and then generates questions to get response that matches the objective and sub-objectives. Further, a sequence is decided for the questions, and accordingly recommendations are generated. The system also assesses whether response obtained from users has reached saturation, and accordingly generates trigger to terminate data collection.

Abstract: An entity resolution data structure system compares two data graphs by creating a confusion matrix in a distributed processing environment. A benchmark file is created from one data graph for comparison to a reference data graph. Identifiers and metadata are appended to the benchmark file to allow the comparison to take place and the construction of a confusion matrix. The confusion matrix provides a high-level indication of the results of the comparison. When the data graphs contain personally identifiable information (Pll), the process does not require any Pll to be transmitted in either direction between the parties who maintain the data graphs to be compared.

Abstract: Semantic type-ahead suggestions are provided for natural language database searches. The system maintains an index of previous natural language database searches and a sorted prefix map based on words used in previous natural language database searches. The system receives user input for a new search, creates a search list based on the user input and user input synonyms, searches the index using the search list, and creates a candidate match list with matching previous searches in the index. If the system determines that the user input includes a database object reference, it filters out searches in the candidate match list that are associated with a different database object. The system ranks the remaining searches in the candidate match list and displays the top n-ranked previous searches in the candidate match list as type-ahead suggestions. The system repeats the steps for each incremental input character and updates the type-ahead suggestions accordingly.

Abstract: An electronic social network can be provided that unifies products and services of one or more entities, such as financial institutions. Users of the social network can provide input regarding one or more products or services. Feedback can be solicited from other users of the social network regarding the input, and a score can be generated for a user that represents a level of agreement of the other users. A recommendation is generated and conveyed to users of the social network based on the score and user profile of the user associated with the score.

Abstract: A systems including one or more processors and one or more non-transitory computer readable media storing computing instructions that, when executed on the one or more processors, perform: receiving in-session user activity information, the in-session user activity information comprising a user search query from a user and a ranked list of products to be displayed to the user based on the user search query, and a product of the products being part of a boost category; determining a respective content model prediction score for each of the products of the ranked list of the products; determining a respective discounted cumulative gain (DCG) score for each of the products of the ranked list of the products; analyzing the product that is part of the boost category to determine if the product is to be repositioned within the ranked list of the products based on the respective DCG score for each of the products; and in response to determining that the product that is part of the boost category is to be repositione

Abstract: A method of latent intent clustering is provided. The method comprises encoding identified features in a number of electronic user reports in a database. A binary matrix is created, wherein each row of the binary matric represents a different report and each column represents a different available feature. A 1 is placed in each cell of the matrix that matches a feature present in a user report. Cosine similarities are calculated for the user reports, and a similarity matrix is created, wherein each row and column of the binary matrix represents a different report, and wherein the cosine similarities of the reports are placed in corresponding cells of the matrix. The reports are clustered according to the cosine similarities. Features common to reports in each cluster are identified, and an intent of each report cluster is labeled according to the common features.

Abstract: Techniques for suggesting accessory devices controlled by an application executing on a mobile device are disclosed. A method includes measuring one or more sensor values using one or more sensors of a mobile device and the one or more sensor values are determined from one or more signals emitted by a first one or more accessory devices. An area of a physical space for the first one or more accessory devices can be determined based on the one or more sensor values. A second one or more accessory devices associated with the same area as the first one or more accessory devices can be suggested to a user.

Abstract: A system to generate and present an intelligent RSS feed that performs operations that include: accessing a media collection from a media host, wherein the media collection may comprise a podcast; caching a sequence of media objects associated with the media collection at a media repository; receiving a request that comprises an identifier associated with the media collection; inserting a media object at a position among the sequence of media objects; and causing display of a presentation of the sequence of media objects at a client device wherein the sequence of media objects includes one or more podcasts, and the presentation of the sequence of media objects includes the media object inserted at the position among the sequence of media objects.

Abstract: A system located in a building. The system including a processing circuit configured to receive tags describing points of a piece of building equipment, the piece of building equipment connected to the system. The processing circuit configured to map the tags to classes of a schema of a graph data structure. The processing circuit configured to perform clustering to generate clusters of the points. The processing circuit configured to identify, based on the clusters, relationships in the schema of the graph data structure between the tags mapped to the classes of the schema of the graph data structure. The processing circuit configured to communicate data to a second system based at least in part on the tags mapped to the classes and the relationships.

Abstract: Techniques regarding pipeline security compliance are provided. For example, one or more embodiments described herein can comprise a system, which can comprise a memory that can store computer executable components. The system can also comprise a processor, operably coupled to the memory, and that can execute the computer executable components stored in the memory. The computer executable components can comprise a certification component that certifies a result from an execution of a pipeline for a computer application based on a comparison of an implemented security control profile associated with the pipeline and an applicable security control profile identified from a source artifact associated with the computer application.

Abstract: Disclosed are some implementations of systems, apparatus, methods and computer program products for integrating question generation and answer retrieval in a question answer system. The system generates a question using a set of documents and determines whether it is semantically distinct from questions in a question-answer repository. After determining that the question is semantically distinct from questions in the question-answer repository, the system adds the question to the question-answer repository. Upon receipt of a user-submitted question, the system uses the question-answer repository to identify a semantically similar question. The system retrieves an answer corresponding to the identified question from the question-answer repository and provides the answer in response to the user-submitted question.

Abstract: Iterative search of a candidate resume database is performed using computer-implemented methods and accompanying systems, and includes the steps of generating a ranked list of search terms associated with a job candidate search, constructing a search query of the candidate resume database based on the ranked list of search terms, executing the search query on the candidate resume database in a count mode to generate a number of results, evaluating the number of results generated from the execution of the search query by comparing the number of results to individual thresholds within a set of thresholds. Upon determining that the number of results exceeds one or more thresholds, the methods further include retrieving the results from the candidate resume database. Upon determining that the number of results does not exceed one or more thresholds, the methods further include removing a restrictive term from the search query.

Abstract: A data analysis system for measuring a materiality feature of interest is disclosed. The system includes a computing cluster ingesting content comprising a plurality of observables relevant to an entity, wherein each observable is related to at least one feature of interest. The system further includes an extraction engine running on the computing cluster and tagging the observables with an entity identifier in response to the observables referencing at least one of an entity, a tradename associated with the entity, or product associated with the entity. Additionally, the system includes an analysis engine running on the computing cluster and tagging an observable in response to the feature of interest being related to the observable. In one embedment, the analysis engine measures the materiality of the feature of interest to the entity by counting a number of observables from the plurality of observables tagged with the entity identifier.

Abstract: The disclosed technologies address problems that users face when using email applications. Various embodiments generally relate to systems and methods for organizing and prioritizing emails in an email application so that a user has greater control over a display of email messages.

Abstract: An information processing apparatus includes at least one processor, in which the processor receives a search objective target selected as a search objective from among a plurality of objective targets which are objective target subjects or objective target objects capable of being the search objective, receives a search condition designated for the search objective target, outputs a search result obtained by searching for a search target based on the search condition, and accumulates the search condition used for a search performed as the search objective target, as a search history, for each of the plurality of objective targets.

Abstract: Disclosed is an approach for evaluating internal consistency of object classifications using machine learning modeling. In an example, models are iteratively trained, using supervised learning, on different majority segments (e.g., about 90%) of a dataset as training data segments. Trained models can be applied to the remaining data (e.g., about 10%) as test data segments to obtain, for each object, a predicted classification and a confidence score. Models in training iterations (e.g., 10 iterations) may be independently trained on substantially non-overlapping test data segments (with each iteration testing, e.g., about 10% of the dataset). When a model's predicted classification disagrees from a prior classification, and the confidence of the prediction is sufficiently high (indicating sufficiently strong disagreement), the object's prior classification may be revised. Training data, other than the data itself being evaluated for consistency, is not necessarily required.

Abstract: A method of retrieving a query, an electronic device, and a medium are provided, which relate to fields of intelligent searching, intelligent recommendation, natural language processing, etc. The method of retrieving a query includes: processing a query to be retrieved, so as to obtain a first processing result for the query to be retrieved, the first processing result containing a first text information and a first semantic information; comparing the first processing result with a second processing result of each of at least one candidate query, so as to obtain a comparison result, the second processing result containing a second text information and a second semantic information; determining a target query from the at least one candidate query based on the comparison result; and determining at least one of the target query or an answer corresponding to the target query as a retrieving result.

Abstract: A device may receive input data associated with a legal regulation, and may process the input data to generate a record that includes: the input data in a knowledge representation format and a semantic representation format, data identifying a feature, data identifying an industry classification, or data identifying an entity of interest. The device may process the record, with machine learning models, to determine output data that includes: data indicating that the legal regulation is inconsistent, data indicating that the legal regulation is outdated, data indicating a sentiment for the legal regulation, data indicating a prescriptive nature of the legal regulation, data indicating a complexity of the legal regulation, data indicating a misrepresentation in the legal regulation, data indicating a compliance burden associated with the legal regulation, or data indicating an industry performance impact of the legal regulation. The device may perform actions based on the output data.

Abstract: In some embodiments, a computing system computes a hierarchical entity data model to facilitate autocompleting forms by generating an electronic schema extraction from an electronic form lacking data for one or more fields. The computing system generates an electronic schema including an input category and input field elements. The computing system accesses a hierarchical entity-data model including and entity category and entity-data elements. The computing system identifies associations between the entity category and input category based on semantic matching including text of an entity category label and an input field category label or matching a number of fields within an entity category to an input category. The computing system verifies the association by applying a natural language processing engine to the input field elements and the entity-data elements. The computing system autocompletes one or more input field elements with entity data from one or more of the entity-data elements.

Abstract: Systems and methods for debiasing a recommendation engine are disclosed herein. A search query associated with a user profile is received at a recommendation engine. Control circuitry generates a result set of items of content based on the search query and generates a bias score for a content attribute based on the result set. The control circuitry also generates a time-averaged bias score for the content attribute based on a plurality of search queries associated with the user profile. Based on the bias score and the time-averaged bias score, the control circuitry determines whether a bias is signaled for the content attribute. Finally, the control circuitry outputs, for display via a computing device, the result set or a debiased result set based on a result of the determination of whether the bias is signaled.

Abstract: A method for maintaining security in a distributed augmented-reality (AR) system is provided. The method may include generating a database that includes a list of objects associated with sensitive information. The method may include receiving requests from a plurality of user devices to participate in an AR session, and determining an authorization level to associate with each of the plurality of user devices. The method may include receiving, from a first one of the plurality of user devices, an image that was scanned in via a camera of the first user device. The method may include processing the image to determine if the image includes one of the objects. When the image includes one of the objects, the method may include masking the object in the image.

Abstract: The disclosed technology provides for generating a lossless representation of a webpage on a client device. A method can include rendering, by a client device, a webpage to be presented in a web browser at the client device, receiving an indication of user input to run a content check on the rendered webpage, calling, in response to receiving the indication of user input, a scraping function to scrape a document object model (“DOM”) of the rendered webpage in the web browser at the client device, generating a lossless representation of the scraped DOM of the rendered webpage, and transmitting, to a web content analysis system, the lossless representation of the scraped DOM to be used when running the content check for the rendered webpage. The lossless representation of the scraped DOM can be a flat object model (“FlatOM”) of the DOM of the rendered webpage.

Abstract: Systems, computer program products, and computer-implemented methods for generating interactive graphical user interfaces, software-based workflows, and data integrations using catalogs of workflow applications and auto-generation of aspects of the workflows.

Abstract: This application provides a keyword extraction method. The method includes: receiving an information entity including a title and a text; performing word segmentation on the text, to obtain a plurality of candidate words; and performing character segmentation on the title corresponding to semantics of the text, to obtain a plurality of characters; sequentially inputting the plurality of candidate words to a keyword extraction model, to obtain an attention weight of each candidate word relative to each character; selecting, from the plurality of candidate words, a candidate word that appears in the title; determining an extraction threshold according to an attention weight of the selected candidate word relative to each character; and determining a keyword of the text of the information entity from the candidate words according to the extraction threshold. This application further provide a method for training a keyword extraction model, a computer device, and a storage medium.

Abstract: The present disclosure is directed towards systems and methods for finding documents that are similar to a reference text. The inventive systems and methods examine a set of collected documents to determine the facts present in those documents by, for example, extracting triplets and expanding them. A user's input reference text is similarly examined to extract and expand triplets therein and the facts identified with respect to the reference text are used as a basis to find documents having similar facts. The present disclosure is also related to systems and methods for mining facts from documents relating to a primary source such as a piece of legislation and using the mined facts to improve the results of subsequent searches.

Abstract: A system and method for updating computerized language models is provided that automatically adds or deletes terms from the language model to capture trending events or products, while maximizing computer efficiencies by deleting terms that are no longer trending and use of knowledge bases, machine learning model training and evaluation corpora, analysis tools and databases.

Abstract: Embodiments of the present disclosure provide a data gathering and query method for collecting ongoing updates to large, unstructured or semi-structured databases. The method comprises gathering a plurality of events defined in a database syntax that is not structured and aggregating the plurality of events into one or more part files. Each of the one or more part files store a subset of the plurality of events in a columnar format, and each of the one or more part files comprises a header file that includes metadata corresponding to a subset of the plurality of events stored in the part file and is separate from the subset of events stored in the part file. The method further comprises uploading the one or more part files to a cloud storage repository configured to store the one or more part files so that they can be queried by a query server based on the header files.

Abstract: Provided are a method and apparatus for processing map information, a device, and a storage medium which relate to the field of computer technology and, in particular, to the fields of intelligent transportation and computer vision technology. The specific implementation includes that an object in an image for query is recognized to obtain an object recognition result and that a target point of interest matching the image for query is selected from at least one candidate point of interest of an electronic map according to the object recognition result.

Abstract: Implementations generally relate to evaluation of similar content-based images. In some implementations, a method includes receiving a first image, where the first image includes at least one first object. The method further includes receiving a second image, where the second image includes at least one second object. The method further includes computing a structural similarity index measure (SSIM) value based on the at least one first object and the at least one second object. The method further includes computing a scale invariant feature transform (SIFT) value based on the at least one first object and the at least one second object. The method further includes computing a histogram value based on the at least one first object and the at least one second object. The method further includes computing a similarity score based on the SSIM value, the SIFT value, and the histogram value.

Abstract: Techniques performed by a data processing system for diagnosing problems with a communications platform include obtaining query parameters including an aggregation operator for invoking a machine learning algorithm configured to analyze performance data for the communications platform, automatically executing the query on the performance data to obtain query results by invoking the machine learning algorithm on the performance data to automatically identify a plurality of clusters of data indicative of a performance problem, and presenting a visualization of the query results. The visualization includes indicators identifying cluster properties for which the query results are further refinable and one or more second indicators identifying the second subset of the second cluster properties which are not relevant for further refining the first query results. The indicators are actuatable to automatically update and re-execute the first query based on the respective indicator that is actuated.

Abstract: Described herein are mechanisms to allow users to access functionality of applications in a suite of applications. In a first aspect, when a query relating to functionality of an application is received from a user, an index containing both top-level and sub-level functionality is searched. Results are ranked using a trained machine learning model using both context describing user interactions and the search results. A subset of the ranked results are presented to the user as options. In a second aspect the index can comprise entries describing functionality from other applications so that results presented to the user can include cross-application functionality. In a third aspect, the index can be searched using the context prior to receiving a query and adjusting the user interface based on the results. In a fourth aspect, the system can recommend other applications and/or devices that are better suited to a user's intent.

Abstract: Aspects described herein may allow for generating a customized price rating using a machine learning algorithm. This may have the effect of improving the display of information about merchants by including customized, personalized price ratings that better reflect the tastes and preferences of a user or group of users. According to some aspects, these and other benefits may be achieved by using a machine learning model, trained to receive input corresponding to both user data and merchant data and output an indication of customized price rating(s) for the merchant that is specific to one or more users, and then to generate information about the merchant for display that includes the customized price rating(s).

Abstract: One or more Bluetooth® low energy (BLE) beacons in communication with a remote server that provides check in capabilities and payment capabilities may be installed at a location. The BLE beacons may connect with a user's mobile device when the user enters the location and allow the user to check in to the location and authorize payments to be made at the location. Once the user is checked in to the location, the user may be provided with additional functionality, benefits, offers, and applications related to the location and facilitated by the check in. Further, the user may be pre-checked in into a next location when the user is at a current location.

Abstract: A method of training an image recognition model includes: generating a virtual dynamic vision sensor (DVS) image using a virtual simulator; generating label information including information about a correct answer to a result of recognition of the DVS image by the image recognition model, with respect to the DVS image; and training the image recognition model by modifying the image recognition model so that a difference between the result of recognition of the DVS image by the image recognition model and the label information is reduced.

Abstract: A computer-implemented method, a computer program product, and a computer system for misinformative content mitigation. A computer receives from a social network system a request for curated content, where the curated content is related to misinformative content that has been identified by the social network system. The computer sends to a system of a paid curated content producer a request for making the curated content public. In response to that the system of the paid curated content producer accepts the request for making the curated content public, the computer provides to the social network system a link to the curated content on the system of the paid curated content producer. On the social network system, the misinformative content is flagged and the link to the curated content is presented aside the misinformative content.

Abstract: A client application sends an application programming interface call to an interface of a content management system. The call specifies one or more content item search criteria. Sending the call causes the content management system to perform a lookup in a content item index to identify at least one content item that satisfies the one or more content item search criteria. Based on sending the call, the client application receives from the content management system a suggestion to attach the at least one content item to a text being displayed by at the computing system.

Abstract: Embodiments provide methods, and server systems connecting with a customer service representative/agent for a financial transaction. A method includes sending, by a server system, a message related to a financial transaction to a user device using a messaging protocol. The message includes at least one option to receive a response of the message from the user device for connecting with an agent. The method includes receiving the response of the message from the user device. The response includes an authentication data. The method includes verifying the authentication data. Upon successful verification, the method includes electronically extracting at least one context of the message from the response. The method includes sending the at least one context to a Customer Service Centre (CSC) server. The CSC server is configured to assign the agent to the user based on the at least one context for an upcoming conversation about the message.

Abstract: At least some embodiments are directed to an entity matching detection system. The entity matching detection system includes a latent similarity identification machine learning model that receives one or more data records and generates a final similarity score indicative of a latent similarity between the one or more data records and a second data record. The entity matching detection system can identify lexical and semantic similarities between attribute values and can analyze and compute similarity scores for direct-linked attribute values and cross-linked attribute values extracted from different data records.

Abstract: Methods and apparatuses are described in which unstructured computer text is analyzed for generation of an enforcement action database. A computing device receives a digital document comprising a corpus of unstructured text and converts each sentence into tokens. The computing device identifies entities in the tokens and classifies each sentence as relevant or not relevant to an enforcement action. For each relevant sentence, the computing device generates a record in a first data structure and generates a record in a second data structure. The computing device maps the records in the first data structure to the records in the second data structure based upon an enforcement attribute to generate a third data structure. The computing device aggregates the records in the third data structure based upon a name of the entity and a type of the entity in the sentence to determine an aggregated first enforcement attribute for each entity.

Abstract: A client application sends an application programming interface call to an interface of a content management system. The call specifies one or more content item search criteria. Sending the call causes the content management system to perform a lookup in a content item index to identify at least one content item that satisfies the one or more content item search criteria. Based on sending the call, the client application receives from the content management system a suggestion to attach the at least one content item to a text being displayed by at the computing system.

Abstract: Computer implemented methods for executing a database query in a database using an electronic computing device are presented, the method including: causing the electronic computing device to provide at least one data structure including at least one object having at least two properties of different data types; identifying a property of the at least one data structure that includes the smallest unique values in terms of type of data stored in that particular property; executing the database query, including any limiting parameters, configured to retrieve only data from the property including the smallest unique values that arc retrievable for a record; retrieving a predefined number of results of the executed database query; and for each set of predefined number of retrieved results, executing a new database access thread that is configured to retrieve data that are present in the records identified with the smallest unique values.

Abstract: A method of generating and optimizing a codebooks for document analysis comprises: receiving a first set of document images; extracting a plurality of keypoint regions from each document image of the first set of document images; calculating local descriptors for each keypoint region of the extracted keypoint regions; clustering the local descriptors such that each center of a cluster of local descriptors corresponds to a respective visual word; generating a codebook containing a set of visual words; and optimizing the codebook by maximizing mutual information (MI) between a target field of a second set of document images and at least one visual word of the set of visual words.