Abstract: Aspects described herein may allow for the application of machine learning techniques to the classification of transactions and/or the prediction of shipping means for such transactions. This may have the effect of generating better insights into user shopping behavior and providing users with better predictions as to how they may expect to receive their purchases. A browser application may be monitored to determine that a user is shopping. A machine learning model may be used to predict shipping data associated with a product and/or merchant associated with the shopping of the user. Predicted shipping data may be displayed to the user, and user feedback may be requested and received to confirm model predictions and review purchases and shipping experiences.

Abstract: An information search system includes: a database configured to store service support information; and an information management device configured to search for predetermined information from the service support information, the information management device including a processor configured to receive an input of a query including a keyword, from a searcher, receive an input of classification information to which a narrowing keyword belongs, present a narrowing keyword belonging to the input classification information and having a correlative relation with the keyword, in a selectable manner, and execute a searching process across the database, based on the input keyword and the selected narrowing keyword.

Abstract: An example includes one or more devices may include one or more memories and one or more processors, communicatively coupled with at least one of the one or more memories, to identify a service that is provided within a region; identify a model that is associated with the service, the model having been trained based on consumer profile data, service provider data, and historical information; determine a current demand associated with the service in the region; predict, using the model and based on the current demand associated with the service, a future demand for the service during a time period; determine a current capacity to provide the service based on real-time service provider information associated with service providers that are providing the service in the region; and perform an action associated with the service based on the future demand for the service and the current capacity to provide the service.

Abstract: Systems and methods are provided for creating and utilizing search indexes with different properties to provide efficient and flexible means for retrieving and ranking high-dimensional content. A first index and a second index are generated with a different sets of properties, which may include a permutation prefix length and a vector quantization scheme, and are each associated with a computational workload. Further, at least one index structure may be created with a posting list associated with a permutation prefix and a set of references, rather than only a permutation prefix, to balance out posting lists of variable lengths within the index. Another embodiment of the disclosure provides for expansion of search results at search time, rather than during indexing, by retrieving posting lists associated with shorter permutation prefix after an initial list is determined to have insufficient recall.

Abstract: An online system may generate numerous search records in response to searches requested by users. The online system may use a specific way to sample the historical search records to reduce biases in sampling. For example, the online system retrieves historical query records associated with an item query engine. The set of historical query records includes a plurality of search phrases. A historical query record is associated with a search phrase and a list of items returned by the item query engine. The online system determines the search frequencies for the search phrases. The online system stratifies the historical query records into a plurality of bins according to the search frequencies of the search phrases. The online system samples the historical query records from the plurality of bins to collect a representative set of historical query records and outputs the representative set of historical query records for rating.

Abstract: Methods, systems, and computer program products for auto-discovery of reasoning knowledge graphs in supply chains are provided herein. A computer-implemented method includes obtaining a spatiotemporal query related to a demand of at least one product in a supply chain; analyzing the spatiotemporal query to identify one or more parameters affecting the demand of the at least one product, wherein the one or more parameters comprise at least one of one or more climate parameters and one or more disruptive event parameters; generating a knowledge graph comprising information indicating an impact on the demand of the at least one product for at least a portion of the one or more parameters; and outputting, to a user interface, an explanation of a predicted demand forecast for the at least one product based at least in part on the knowledge graph.

Abstract: An automatically detecting method for time-varying text region of interest is disclosed. The automatically detecting method is adapted to an image processing unit of an information retrieval system, to detect a time-varying text region of interest having specific characters or character set as unit on an operation screen of a manufacturing machine, a processing machine or other equipment; furthermore, the automatically detecting method can be performed based on presence or absence of the historical screen data, and union of the detected region proposals for the time-varying text region of interest, to obtain an automatically labeled and selected time-varying text region of interest. According to the automatically detecting method, the user only needs to confirm whether the required data are labeled and selected, so it is more convenient for the user to setting data, and greatly helpful to reduce the setting time and correctly detect the required information.

Abstract: Techniques for expense report submission are disclosed. An expense report submission system receives, via a graphical user interface, a user query that corresponds to requesting whether a particular expense is allowed. The expense report submission system applies the user query to a machine learning model configured to evaluate data associated with expenses against one or more expense policy rules. The expense report submission system generates a response to the user query based at least on a result of applying the user query to the machine learning model. The response to the user query indicates whether the expense is allowed based at least on the expense reporting rule(s). The expense report submission system presents, in the graphical user interface, the response to the user query.

Abstract: A method of training a machine learning model to determine an item margin is provided. The method includes monitoring a first value for a first item having attributes and monitoring a first value for a second type of item having attributes where an attribute of the first attributes is the same as an attribute of the second attributes. The method also includes determining a first margin based on the first values. The first attributes, the second attributes, and the first margin are input as training data for the machine learning model where the machine learning model is trained with the training data. The monitoring operations for the first item and the second item are repeated to obtain a second value for the first and second items. Furthermore, the trained machine learning model is applied to the second values to determine a second margin.

Abstract: In one embodiment, the present invention includes a computer-implemented method comprising storing data in an application using an application custom data type and application custom data structure. The data is stored in a database using the application custom data type and the application custom data structure. In one embodiment, a request is sent to access the data from the application to the database. The data is retrieved from the database in response to the request in the application custom data type and the application custom data structure. In one embodiment, the data is sent from the database to a shared memory in the application custom data type and the application custom data structure and the data is retrieved by the application from the shared memory in the application custom data type and the application custom data structure.

Abstract: Distributed pseudo-random subset generation includes obtaining a data-query indicating a first table having a first column including unique values, a second table having a second column including unique values, a join clause joining the first table and the second table on the first column and the second column, and a limit value, pseudo-random filtering the first table to obtain left intermediate data and left filtering criteria, pseudo-random filtering the second table to obtain right intermediate data and right filtering criteria, obtaining intermediate results data by full outer joining the left intermediate data and the right intermediate data, obtaining results data by filtering the intermediate results data using most-restrictive filtering criteria among the left filtering criteria and the right filtering criteria, and outputting the results data, wherein outputting the results data includes limiting the cardinality of rows of the results data to be at most the limit value.

Abstract: In one embodiment, the present invention includes a computer-implemented method comprising storing data in an application using an application custom data type and application custom data structure. The data is stored in a database using the application custom data type and the application custom data structure. In one embodiment, a request is sent to access the data from the application to the database. The data is retrieved from the database in response to the request in the application custom data type and the application custom data structure. In one embodiment, the data is sent from the database to a shared memory in the application custom data type and the application custom data structure and the data is retrieved by the application from the shared memory in the application custom data type and the application custom data structure.

Abstract: Included are embodiments for determining sort order. At least one embodiment of a method includes determining a plurality of search results associated with at least one search criteria. Some embodiments of a method include determining a first distance between at least one of the search results and a home page associated with the at least one search result and determining a display order for the plurality of search results.

Abstract: Attributes in data records are identified. Episodes corresponding to the respective attributes include respective data records. The episodes are scored, where scoring of a particular one of the episodes is based on relative time densities between successive data records of the particular episode. A visualization of at least some of the episodes to provide an alert of time-dense episodes.

Abstract: A method, process, and associated systems for ranking elements of an answer set retrieved by a conceptual search engine in response to a set of queried concepts submitted as a conceptual query. A processor of a computer system orders by relevance the semantic schemas that make up the retrieved answer set by first identifying a weight of each concept in a retrieved schema as a function of relationships among the concepts in that schema, then by identifying a semantic value of each queried concept in the retrieved schema as a function of the weights, and finally by identifying a semantic value of the query in the schema as a function of the semantic values of all queried concepts. The relative relevance of each schema may then be identified as a function of the semantic value of the query in that schema.

Abstract: A method, process, and associated systems for ranking elements of an answer set retrieved by a conceptual search engine in response to a set of queried concepts submitted as a conceptual query. A processor of a computer system orders by relevance the semantic schemas that make up the retrieved answer set by first identifying a weight of each concept in a retrieved schema as a function of relationships among the concepts in that schema, then by identifying a semantic value of each queried concept in the retrieved schema as a function of the weights, and finally by identifying a semantic value of the query in the schema as a function of the semantic values of all queried concepts. The relative relevance of each schema may then be identified as a function of the semantic value of the query in that schema.

Abstract: Mechanisms are provided to allow for content discovery using closed caption content search and ranking. Search mechanisms analyze titles, descriptions, social media content, metadata, etc., and intelligently organize content for presentation to a viewer. Image recognition and audio recognition algorithms can also be performed to further identify entities or validate results from the analysis of metadata. Other closed captioning content may be analyzed to determine the relevance of a piece of media content to a particular search term found in the piece of media content. Results are ranked based on the prominence of search and related terms in titles, descriptions, and closed caption contents along with the popularity of the media content itself.

Abstract: A task-oriented search tool searches for one or more tasks related to a search term submitted via a computer-user interface. The task-oriented search tool displays search results that include a task related to the search term and a task-solving tool for performing the task related to the search term.

Abstract: A method performed by one or more processing devices includes receiving a request for social media content that is relevant to one or more search terms; identifying, based on a social graph of a user, social media content with connections in the social graph to the user, wherein the social media content is associated with metadata; identifying metadata that is associated with the identified social media content and that has at least a threshold amount of relevance to the one or more search terms; generating a summarization of contents of the metadata identified and contents of the social media content identified; and transmitting, to the device used by the user who sent the request, the summarization for presentation to the user, in response to the request for social media content.

Abstract: A system and method is disclosed for organizing search results in response to a search query. A search query and sorting criteria are received from a web browser, the sorting criteria including a coarse level of granularity and a fine level of granularity. Results are received from a search engine based on the search query and the search results are organized based on the multiple sorting criteria. In this regard, the search results are ordered top-down from the coarse level of granularity to the fine level of a granularity. The organized search results are then provided for display to the web browser.

Abstract: Provided are a computer program product, system, and method for incorporating related searches by other users in a social network in a search request. A search request is received including keywords from a searching user including a keyword to search. A determination is made of at least one other user in a trust relationship with the searching user. A determination is made of related searches submitted by the determined at least one other user having related keywords related to the keywords included in the received search request. Information is returned to the searching user on the related keywords in the related searches in response to determining that the determined at least one other submitted related searches.

Abstract: A system includes a memory operable to store visualization rules. The system also includes a processor communicatively coupled to the memory. The processor is operable to receive a visualization request relating to information stored in an ontology. The processor is further operable to parse the visualization request to generate a search query. The processor is further operable to submit the search query to the ontology. The processor is further operable to receive, in response to the query, a result. The result includes a plurality of instances and a plurality of relationships between the instances. The processor is further operable to generate a visual representation of the result using the visualization rules.

Abstract: Embodiments are provided for displaying ontology data. An ontology is displayed for a selected item. The ontology presents representational elements associated with the selected item, a combination of the representational elements defining and/or establishing a context for the selected item. The representational elements are categorized on a hierarchical path, the elements ordered by scope size. The ontology can be navigated in multiple directions with the directions including items broader in scope, and items narrower in scope. Navigating in one of these directions includes a displayed representational element from the ontology having a scope size reflecting the navigated direction.

Abstract: A system and method of query optimization are disclosed. A complexity measure of a client query received from a client application is compared to a predetermined threshold, which is determined using at least one of system data and query execution data. An optimized query is generated by modifying the client query to reduce the complexity measure of the client query if the complexity measure of the client query exceeds the predetermined threshold. The optimized query is submitted to a server application on a server and a post-processing routine to be applied to the result set is selected.

Abstract: A method, process, and associated systems for ranking elements of an answer set retrieved by a conceptual search engine in response to a set of queried concepts submitted as a conceptual query. A processor of a computer system orders by relevance the semantic schemas that make up the retrieved answer set by first identifying a weight of each concept in a retrieved schema as a function of relationships among the concepts in that schema, then by identifying a semantic value of each queried concept in the retrieved schema as a function of the weights, and finally by identifying a semantic value of the query in the schema as a function of the semantic values of all queried concepts. The relative relevance of each schema may then be identified as a function of the semantic value of the query in that schema.

Abstract: A method for defining a relevant area for displaying advertisements based on a category of a user search is disclosed. The method includes receiving a search request from a user; determining a business category based on the search request; determining a location associated with the search request; determining business category density information based at least on the determined business category and the location associated with the search request; determining an extent based on business category density information for the location; determining local content to provide to the user based at least in part on the determined extent, the business category, and the location; and providing for display to the user the local content, wherein the local content is provided to the user in addition to the search results. The location may be explicitly included in the search request, associated with responsive search results, or associated with the search request.

Abstract: Techniques are described for performing grouping and aggregation operations. In an embodiment, a request is received to aggregate data grouped by a first column. In response to receiving the request, values are loaded from the first column into an input cache. The values include values, from the first column, from a set of rows. A filter unit is programmed with logic to perform a comparison between a particular value, from the first column of a first row, and values in the first column of a plurality of rows, of the set of rows. Based on the comparison, a predicate result is generated that identifies rows, within the plurality of rows, that have a valued in the first column that matches the particular value. An aggregate value for a second column is generated by aggregating values, from the second column, of each of the rows identified by the predicate result.

Abstract: A method, system, and computer readable storage medium is provided for identifying textual terms in response to a visual query is provided. A server system receives a visual query from a client system. The visual query is responded to as follows. A set of image feature values for the visual query is generated. The set of image feature values is mapped to a plurality of textual terms, including a weight for each of the textual terms in the plurality of textual terms. The textual terms are ranked in accordance with the weights of the textual terms. Then, in accordance with the ranking the textual terms, one or more of the ranked textual terms are sent to the client system.

Abstract: Methods and systems are provided for ranking search results and generating a presentation. In some implementations, a search system generates a presentation based on a search query. In some implementations, a search system ranks search results based on data stored in a knowledge graph. In some implementations, a search system identifies a modifying concept such as a superlative in a received search query, and determines ranking properties based on the modifying concept.

Abstract: A data modeling platform allows a user to enter a search request using natural language (e.g., typed or spoken). The search request may be semantically parsed to provide mapping to various database search elements. Pre-existing data models may be identified from the database search elements. One or more new data models may be created from the database search elements. The results represent candidate data models in response to the user's search request.

Abstract: Attributes in data records are identified. Episodes corresponding to the respective attributes include respective data records. The episodes are scored, where scoring of a particular one of the episodes is based on relative time densities between successive data records of the particular episode. A visualization of at least some of the episodes to provide an alert of time-dense episodes.

Abstract: Embodiments improve the ability of users of a social networking system to search for information that is likely to be relevant to them by learning and/or applying a search context associated with selector components used to search for objects of the social networking system. The search context is specific to the use of an individual selector and thus need not be as general as the context of an entire page or set of pages in which selectors can be embedded. The social networking system may learn the context of a selector by monitoring user selections from prior search results performed using the selector.

Abstract: For categorizing content, a categorizing module categorizes content of a content stream into a category of a plurality of categories. A recording module records a navigation action directed to first content of the content stream. The recording module further elevates a ranking of a first category associated with the first content in response to the navigation action. In addition, the recording module displays highest ranking categories for selection.

Abstract: A method, system and a computer program for identifying unexplored research avenues in a plurality of publications is provided. Citation maps for the plurality of publications are generated. The initial set of publications is filtered on the basis of the citation maps and resulting set of publications are ranked according to their prestige value. Natural language processing means are used to perform context matching in order to identify set of sentences in the publications. Paragraphs containing the set of sentences are displayed to a user along with pointers to the respective publication.

Abstract: Method, system, and programs for presenting query results are disclosed. An input is received from a user related to a search query. A first set of query results and a second set of query results are then obtained based on the search query. A query intent is detected based on the search query. A first presentation style associated with the detected query intent is determined for the first set of query results. The first and second sets of query results are provided to the user. The first set of query results are presented in the first presentation style.

Abstract: The hierarchal clustering method for large XML data is a scalable technique that takes large XML data sets and measures distances between those sets of data and then arranges those data into groups according to the distance measurements. The method can cluster XML data by both content and structure, and can cluster homogeneous XML data.

Abstract: Improved, personalized searching through the use of personal connection graphs is presented. Responsive to receiving a search query, the search engine obtains a user sketch (representing a user's connection graph) and obtains search results responsive to the search query and also based, in part, as a function of a relationship distance between the user and the referenced content as determined by the user sketch. One or more search results pages are generated from the obtained search results and returned to the user.

Abstract: Techniques for annotating an entity in a document corpus using cross-document signals. A method includes determining which documents in a document corpus mention an entity of interest, clustering the documents that mention an entity of interest according to a temporal signal, a structural signal and/or a content signal, thereby forming at least one cluster of documents, and annotating at least one document in the at least one cluster of documents by marking each occurrence of the entity in the at least one document.

Abstract: For filtering web search results, a method includes extracting metadata attributes and associated attribute values from web search results from a web search engine. The web search results organized into a results list with web page data grouped as an entry in the results list. The metadata attributes and associated attribute values extracted from the results list. The method includes presenting the extracted metadata attributes and receiving input from the user indicating one or more selected metadata attributes and a position indication for each selected metadata attribute. Each position indication indicates where in a custom report that attribute values for each selected metadata attribute are to appear. The method includes filtering the received results list based on the selected metadata attributes and displaying the filtered results list to the user in a custom report arranged by the selected position indication for each selected metadata attribute.

Abstract: Systems and methods are provided for aggregating a local media collection of a user with one or more shared media collections shared with the user to provide an aggregate media collection for the user. In general, one or more shared media collections shared with the user are first discovered. Optionally, the user may select one or more of the shared media collections for aggregation with the local media collection of the user. The shared media collections selected by the user, or alternatively all of the shared media collections, are then aggregated with the local media collection of the user to provide an aggregate media collection for the user. The aggregate media collection of the user may then be pruned or filtered based on user preferences of the user.

Abstract: Systems, software, and computer-implemented methods can be used to present search results and search refinement suggestions identified in response to a search query using a search refinement wheel. In response to receiving a search query, a set of suggested query refinements are obtained. A spoke graph structure processable to present a visual representation of the search query and at least a portion of the set of the at least one search query refinements is generated, where the search query is represented by a first node in the visual representation and each of the portion of the set of the search query refinements is represented by an individual spoke in the visual representation that can be activated to generate a corresponding new search query. The generated graph can be provided in response to the search query, e.g., to a client device from which the search query is received.

Abstract: A computer-implemented method for performing a semantically enriched search of services includes: receiving a search string that a user inputs for searching services in a repository; generating queries from the search string; searching a multi-document index using the generated queries, the multi-document index including, for each of the services, an index entry comprising documents interlinked with each other, each of the documents reflecting at least one aspect regarding the service; and presenting an outcome of the search to the user in response to receiving the search string.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for providing search result page previews. In one aspect, a method includes receiving data that specify a set of search results responsive to a search query. Query-relevant content is selected to be included in a page preview for at least one of the search results. In turn, data that cause presentation of the page preview are provided. The data provided can cause presentation of the query-relevant content at an initial zoom level and at a higher zoom level, where the initial zoom level is a zoom level at which both the query-relevant content and other content from the resource are presented. The page preview can include a page tear that defines multiple portions of the page preview for a resource.

Abstract: Systems and methods for responding to search queries with search results and related pivoted search queries is presented. Responsive to a search query, one or more entities related to the search query are identified as pivot entities. For each pivot entity a set of pivoted search queries is identified and included among the search results responsive to the search query.

Abstract: Additional date term granularity may be added to query results and query interfaces used to append query results. Initially, a partial execution of a query may be triggered to identify fields and field types. Fields may be classified as date fields based on their field type. The query may be appended to compute one or more time periods for each classified date field. Each time period may identify a broader time period than the data included in a corresponding classified date field. A query result, which may include the computed time period(s) and time period field(s), may be generated from the execution of the appended query.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer-readable storage medium, and including a method for providing the search results. The method comprises receiving a query from a client device. The method further comprises responsive to the query, identifying, using one or more processors, search results including one or more resources. The method further comprises for at least one resource of search results, determining, using the one or more processors, a size of a data transfer required to access the one resource. The method further comprises providing the search results to the client device including providing a label associated with the one resource indicative of the size.

Abstract: Methods and arrangements for generating exploratory search results. A first search query is accepted in a browsing session, and first search result set is derived, comprising at least one search result from the first search query. A second search query is predicted, and a second search result set is derived, comprising at least one search result from the second search query. A final search result set is yielded via combining the first search result set and the second search result set.

Abstract: Systems and methods of searching by a processor are described herein. In accordance with one aspect of the present disclosure, a master index having a plurality of index documents corresponding to artifacts stored in a repository is generated. An index document maps contents from a source artifact to index categories having different default boost values indicating relative importance of the index categories. A search query is fired by a user to search the master index. A result list of artifacts having contents matching (hit) the search query is generated. Artifacts are ranked in the list based on hit scores according to number of hits, which categories the hits occurred and the default boost values of the index categories.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for presenting a visual search history. Presentation data is received indicating that a resource has been presented by a user device, the resource having been presented in response to user interaction with a search result that referenced the resource. In response to receipt of the presentation data, an image of the resource is acquired. The acquired image is associated with a search query for which the search result was provided. In response to receipt of a request for a search history for the user device, the acquired image and the associated search query are displayed.

Abstract: Extracting principal content from Web pages includes identifying and classifying items on the Web page, building a list of candidates, calculating candidate scores, selecting a top score candidate, performing clean up processing for the top score candidate, and performing final page processing for the top score candidate. Candidate scores may vary according to a number of paragraphs and images grouped according to size. A world length of CJK (Chinese-Japanese-Korean) text may be determined according to punctuation therein. Candidate scores may be modified according to a number of containers and pieces and wherein a container is a Web page element that is associated with tags ‘body’, ‘div’, ‘td’, ‘li’, ‘article/section’ and pieces are candidates that do not include other candidates. Candidate scores may be modified according to a number of ratios corresponding to text and link density.