Abstract: Methods, systems, and apparatus, including computer program products are provided for responding to search queries having results that identify books. In one aspect, a search query and multiple web pages that satisfy the search query and have a ranked order as responses to the search query are received. A subset of web pages that are each a reference page for a respective book are selected. A web page is a reference page for a book when the web page includes a reference to the book and satisfies a citation criterion for the book. A book score is assigned to each of the books for which there is at least one reference page in the group of highest ranking web pages. The book scores are used to select one or more of the books. A book reference is generated for each of the books and the book references are provided in response to the search query.

Abstract: A LPM search engine includes a plurality of exact match (EXM) engines and a moderately sized TCAM. Each EXM engine uses a prefix bitmap scheme that allows the EXM engine to cover multiple consecutive prefix lengths. Thus, instead of covering one prefix length L per EXM engine, the prefix bitmap scheme enables each EXM engine to cover entries having prefix lengths of L, L+1, L+2 and L+3, for example. As a result, fewer EXM engines are potentially underutilized, which effectively reduces quantization loss. Each EXM engine provides a search result with a determined fixed latency when using the prefix bitmap scheme. The results of multiple EXM engines and the moderately sized TCAM are combined to provide a single search result, representative of the longest prefix match. In one embodiment, the LPM search engine supports 32-bit IPv4 (or 128-bit IPv6) search keys, each having associated 15-bit level 3 VPN identification values.

Abstract: A computer system that includes a computer that couples with a database. The computer includes program code or modules to gather location and activity content from disparate sources, and through text analytics, extract associations from the content and populate the database with the associations between locations and activities. Further modules provide end user interaction through presentation of a search user interface specific to locations and activities. Additional modules provide the capability to search the database, rank the results of the search and present the results to the user.

Abstract: A video classification score boosting method boosts classification scores for videos for increased accuracy. A target video is classified with a classifier, producing a classification score. Related video scores are determined using the classifier for sets of videos related to the target video. The sets of related videos may include co-browsed videos, co-commented videos, co-queried videos, and co-uploaded videos. The related video scores may be the mean or median classification score for the classified sets of related videos. Weighting coefficients associated with the classifier are retrieved and applied to the classification score and the related video scores. The weighting coefficients may be determined for the classifier by classifying sets of pre-classified videos with the classifier and determining the weighting coefficients which, when applied to the classification scores of the pre-classified videos, improves the accuracy of the classification scores.

Abstract: The invention provides a display interface in a social networking system that enables the presentation of information related to a user in a timeline or map view. The system accesses information about a user of a social networking system, including both data about the user and social network activities related to the user. The system then selects one or more of these pieces of data and/or activities from a certain time period and gathers them into timeline units based on their relatedness and their relevance to users. These timeline units are ranked by relevance to the user, and are used to generate a timeline or map view for the user containing visual representations of the timeline units organized by location or time. The timeline or map view is then provided to other users of the social networking system that wish to view information about the user.

Abstract: An information processing apparatus according to the present invention includes a data retrieval unit for obtaining at least two element data, a dissimilarity calculation unit for calculating a dissimilarity between the element data obtained by the data retrieval unit, a transition cost calculation unit for calculating a cost of transition from one of the element data obtained by the data retrieval unit to another of the element data thereof which is different therefrom, and a distance calculation unit for calculating an element distance representing the degree of dissimilarity between the element data by using the dissimilarity calculated by the dissimilarity calculation unit and the transition cost calculated by the transition cost calculation unit.

Abstract: A system and method are provided for intelligently, or programmatically, assigning weights for one or more criterion utilized to score media content items based on an analysis of a group of media content items. In general, scoring criteria to be used to score media content items for a user are defined. A group of media content items associated with the user is then analyzed with respect to the criteria to provide results such as a number or percentage of media content items from the group of media content items that satisfy each of the scoring criteria. Based on the results of the analysis, a weight is assigned to each of the scoring criteria. Thereafter, media content items are scored for the user as a function of the weights assigned to the scoring criteria.

Abstract: A method of setting a threshold similarity score value for a first plurality of network user identifiers. The first plurality of network user identifiers, a second plurality of network user identifiers and characteristic data associated with the network user identifiers is received. A performance target and an experimental threshold similarity score value are designated. A similarity score between the first and second plurality of network user identifiers is calculated. Performance statistics data for each of the second plurality of network user identifiers having a similarity score greater than or equal to the experimental threshold similarity score value is collected and compared to the similarity score of the network user identifier. Based on the comparison, the experimental threshold similarity score value is adjusted to a similarity score value that achieves the performance target and the threshold similarity score value is set to the adjusted experimental threshold similarity score value.

Abstract: An approach is provided for generating a compilation of media items. A plurality of media items is received. Respective context vectors for the media items are determined. The context vectors include, at least in part, orientation information, tilt information, altitude information, geo-location information, timing information, or a combination thereof associated with the creation of the respective media items. A compilation of at least a portion of the media items is generated based, at least in part, on the context vectors.

Abstract: The invention discloses a system and method for managing feedback data that will be used for ranking search results. The invention can aggregate a plurality of user feedback data from more than one user into a search index. The user feedback data can be associated with one or more documents within the index such that the one or more documents can be ranked based on the type of feedback data that is aggregated. Once the documents have been ranked, the ranked documents can be provided to a requester.

Abstract: Potential content item slots (e.g., ad slots) in a media (e.g., video, audio, or both) are identified, and each content item slot is associated with a weight that indicates a degree of potential disruption to a flow of the media when a content item (e.g., ad) is inserted in the content item slot.

Abstract: The systems and methods described herein generally relate to increasing user productivity in reviewing query results by visually depicting the presence/absence of a set of query terms in a set of paragraphs across a set of documents.

Abstract: A search engine provides personalized rankings of search results. A user interest profile identifies topics of interest to a user. Each topic is associated with one or more sites, and a boost value, which can be used to augment an information retrieval score of any document from the site. Search results from any search are provided to the user, with a variable control of the ranking of the results. The results can be ranked by their unboosted information retrieval score, thus reflecting no personalization, or by their fully or partially boosted information retrieval scores. This allows the user to selectively control how their interests affect the ranking of the documents.

Abstract: An apparatus (102, 104) and method for generating a formatted clinical study report (206) by analyzing submitted clinical study data (150) based on semantics and maximum entropy analysis at a server (104). Sections of the formatted clinical study report are identified in the submitted data (111) using semantics analysis and maximum entropy analysis (204). The identified sections are formatted and output as the formatted clinical study report (206). The formatted clinical study report (206) may be reviewed and edited after generation by a user interfacing with the server (104).

Abstract: Provided are a method and an apparatus for constructing an ontology for a dialogue system. The method for constructing an ontology for a dialogue system includes: generating a domain ontology plane based on intra-plane relation information of a domain defining a relationship between a plurality of domain nodes; generating a main act ontology plane based on intra-plane relation information of a main act defining a relationship between a plurality of main act nodes; and generating an entity name ontology plane based on intra-plane relation information of an entity name defining a relationship between a plurality of entity name nodes. Therefore, it is possible to construct multiple ontology planes and discriminate components of dialogue frames such as a domain, a main act and an entity name. Also, an effective system response can be performed by discriminating dialogue frames exactly using the multi ontology planes.

Abstract: A method for processing query data is described that includes receiving a query portion from a client over a network. For each of multiple search contexts, a relevance score is determined, based on the query portion. Each search context corresponds to a different set of information against which queries can be executed. Indication of the relevance scores is provided to the client over the network. Determining the relevance score and providing indication are performed prior to an input indicating a complete query or in response thereto. The method may also include associating shortcuts with search contexts, selecting a set of shortcuts based, at least in part, on the relevance scores for the search contexts and the association between the shortcuts and search contexts, and sending the set of shortcuts to the client. The shortcuts include links for accessing a content location associated with the shortcut.

Abstract: Duplicate video search results are detected and removed. Digital signatures are generated for each video content item of a video content corpus. Duplicates are determined for the top n previously received queries by determining the similarity of video content items that are within the same results set of each particular query of the top n previously received queries. Similarities are calculated between any two video documents of the result set of the particular query by measuring the difference between the digital signatures of two video documents. If a similarity between two videos is determined to be above a particular threshold, then the two videos are considered duplicates of each other and the search index is updated by retaining the most relevant of the video documents to the particular query. The less relevant video documents are flagged as duplicates with respect to the particular query.

Abstract: This specification describes technologies relating to displaying online content. In general, one aspect of the subject matter described in this specification can be embodied in methods that include receiving a first query associated with a user request. The methods may further include determining a score for a word in the first query based at least in part on user interaction with a content item served for display in response to a past query that includes the word. The methods may further include selecting a keyword derived from the first query based at least in part on the score. The methods may further include identifying candidate content items using the selected keyword. Other embodiments of this aspect include corresponding systems, apparatus, and computer program products.

Abstract: A method and apparatus to identify names, personalities, titles, and topics that are present in a repository and to identify names, personalities, titles, and topics that are not present in the repository, uses information from external data sources, notably the text used in non-speech, text-based searches, to expand the search terms. The expansion takes place in two forms: (1) finding plausible linguistic variants of existing search terms that are already comprehended in the repository, but that are present under slightly different names; and (2) expanding the existing search term list with items that should be there by virtue of their currency in popular culture, but which for whatever reason have not yet been reflected with content items in the repository.

Abstract: Systems and methods for search and search optimization using a pattern in a location identifier is disclosed. In one aspect, embodiments of the present disclosure include a method, which may be implemented on a system, of search and search optimization. The method includes, detecting a set of location identifiers that have a pattern that matches a specified pattern and identifying a set of search results as having content related to the semantic type. The specified pattern can be stored in a computer-readable storage medium and corresponds to a semantic type. The set of search results can include objects associated with the set of location identifiers having the specified pattern.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for cross-language information retrieval. In general, one aspect of the subject matter described in this specification can be embodied in methods that include the actions of receiving a query in a source language, the query including one or more query terms; automatically determining one or more target languages relevant to the query; translating the query into one or more translated queries in the respective one or more target languages; determining search results responsive to the respective one or more translated queries; and providing one or more of the search results.

Abstract: This application describes methods for searching digital information such as digital documents (e.g., web pages) and computer databases, and specific search techniques such as authority ranking and information retrieval (IR) relevance ranking in keyword searches. In some implementations, the technique includes analyzing digital information viewed as a labeled graph, including nodes and edges, based on a flow of authority among the nodes along the edges, the flow of authority being derived at least in part from different authority transfer rates assigned to the edges based on edge type schema information. In some implementations, the system includes an object rank module configured to generate multiple initial rankings corresponding to multiple query keywords, each of the multiple initial rankings indicating authority of nodes in a graph with respect to each respective query keyword individually; and a query module configured to combine the multiple initial rankings in response to a query.

Abstract: Determining experts based on a search query of a user includes identifying items in a content collection that correspond to the search query, determining authors of the items, and ranking the authors according to relevance to the search query for each of the items for each of the authors. Determining experts based on a search query of a user may also include complementing the query with additional public search results prior to identifying the items. Complementing the query may include using an external data source to search based on the query. The external data source may be selected from the group consisting of Google Search, Yahoo Search, and Microsoft Bing. Determining experts based on a search query of a user may also include presenting the authors to the user in order of ranking The query may be a natural language query.

Abstract: A method of mapping a collection of images, or other higher dimensional items including text documents, and three-and-higher dimensional objects, onto a navigable grid for browsing via a user interface comprises obtaining for each of the images a list of nearest neighbor images and similarity scores for each nearest neighbor; placing a first image on a cell within a grid; from a respective list of nearest neighbors of said first image, finding images that maximize a compatibility score with images already placed on the grid and placing resulting images on neighboring cells; and continuing to place further images on the grid until all cells visible to a user are filled with images. As the user pans or zooms the grid, more cells move into the visible area of the screen and are filled with images in the same way.

Abstract: In a method, system, and computer-readable medium having instructions for semantic matching, a configuration for one or more ontologies is determined with an ontology that has one or more concepts and a representation for the one or more concepts, and the configuration has an assignment of concepts to positions and one or more relationships between concepts in accordance with the representation. The configuration is optimized in accordance with one or more constraints, and a constraint has a relationship defined in a representation for an ontology and a judgment on a similarity of a plurality of concepts from the one or more ontologies, and an estimate is calculated for a similarity between a first concept and a second concept using the configuration.

Abstract: A location classifier generates location information based on textual strings in input text. The location information defines potential geographical relevance of the input text. In determining the location information, the location classifier may receive at least one geo-relevance profile associated with at least one string in the input text, obtain a combined geo-relevance profile for the document from the at least one geo-relevance profile, and determine geographical relevance of the input text based on the combined geo-relevance profile.

Abstract: A computerized method of visualizing the collective opinion of a group regarding one or more qualitative issues. The group initially selects N issues from the universe of potential issues and often assigns the issues images and titles. The system presents each user with graphical user interface screens wherein individual users vote on the relative importance and degree of relationship between the N aspects (Data Points) and issues, often using drag and drop methods. The software computes N×N similarity matrices based on users voting input and clusters various aspects into groups of greater and lesser similarity and importance, and presents results of users qualitative ranking in easy to read relationship tree diagrams where the relative importance and qualitative relationship of the issues may be designated by size and other graphical markers. The software may reside on a network server and present display screens to web browsers running on user's computerized devices.

Abstract: Systems and methods of selecting consumers to receive content on a computer network are provided. A user list identifying a first plurality of users having a group of features corresponding to internet activity of the first plurality of users can be obtained at a computing device. A subgroup of features can be selected from the group of features, and a cluster of users of the first plurality of users can be identified. The users of the cluster of users can each have at least one feature of the subgroup of features. A supplemental user having a supplemental feature can be identified. A correlation between the supplemental feature and at least one feature of the subgroup of features can be determined, and an expanded user list that includes at least one of the first plurality of users and the supplemental user can be generated.

Abstract: Techniques are disclosed for configuring an identity resolution system to support distinct relevance types. Identity records are accessed that are assigned relevance scores of distinct relevance types. Upon determining that the identity records refer to a common individual, the identity records are resolved into an entity representing the common individual. Relevance scores of the distinct relevance types are then determined for the entity, based on the identity records.

Abstract: A method of automatically selecting a number of secondary images and a display template for display with a primary preselected image based on analyzing the primary image's attribute information and comparing the secondary images attribute information and the templates image attribute requirements. The attribute information is used to evaluate a compatibility of the images and template so that a best compatibility fit can be obtained when displaying the images.

Abstract: Methods, systems, and apparatus, including computer program products, are described for reconciling data. In one implementation, a method includes generating co-occurrence scores indicating whether data in entries in a first source of data co-occur within documents in a plurality of documents with data in entries in a second source of data. The co-occurrence scores for a given entry in the first source of data are used to identify a plurality of candidate matching entries in the second source of data for the given entry. Data in fields in the given entry are compared to that of one or more of the candidate matching entries to produce field similarity scores. The field similarity scores and the co-occurrence scores are used to determine a match for the given entry among the plurality of candidate matching entries.

Abstract: An apparatus, system, and method for measuring the similarity of binary objects is disclosed. The method determines at least one pattern signature in an Nth binary object, accessing a location in a similarity store which has object identifiers for each of the previous N?1 binary objects which contain the corresponding pattern, and writing the object identifier of the Nth binary object at that same location in the similarity store. Reporting the number of locations in similarity store which contain the object identifiers of two apparently diverse binary objects is a measure of similarity to each other.

Abstract: Various embodiments are described herein that generally relate to systems and methods for generating context specific terms and performing various actions based on the context specific terms. One example embodiment includes a computer-implemented method for generating context specific terms comprising obtaining a collection of terms from at least one electronic file associated with a given context; comparing the collection of terms with a collection of expected terms to generate candidate terms that are not in the collection of expected terms; determining a relevance for each of the candidate terms; and determining whether to add a given candidate term to a collection of context specific terms for the given context if the relevance for the given candidate term is above a threshold.

Abstract: Aspects of the present invention include object detection training systems and methods and using object detection systems and methods that have been trained. Embodiments presented herein include hybrid learning approaches that combine global classification and local adaptations, which automatically adjust model complexity according to data distribution. Embodiments of the present invention automatically determine model complexity of the local learning algorithm according to the distribution of ambiguous samples. And, embodiments of the local adaptation from global classifier avoid the common under-training problem for local classifier.

Abstract: Techniques are disclosed for configuring an identity resolution system to support distinct relevance types. Identity records are accessed that are assigned relevance scores of distinct relevance types. Upon determining that the identity records refer to a common individual, the identity records are resolved into an entity representing the common individual. Relevance scores of the distinct relevance types are then determined for the entity, based on the identity records.

Abstract: Users collect digital media items such as songs, images, and videos into media libraries. Over time, the user can collect a very large number of media items making organization and use of the media library difficult and time-consuming. The systems and methods described herein alleviate this task by collecting metadata about the media items from multiple sources, determining a similarity between the media items, and clustering the media items with like media items. The systems and methods described herein can position the media items relative to one another in a layout based on their respective similarity. Feedback from the user and from other users can be added to the metadata and used to update the layout of the media items.

Abstract: When a social networking system receives a request from a requesting user for a content item associated with one or more comments, the social networking system determines an interest score for each comment. The interest score for a comment indicates a measure of the user's likelihood of being interested in the comment. Based on the calculated interest scores, the social networking system selects one or more comments for presentation to the viewing user along with the content item. The social networking system may specify an order in which the selected comments are presented based on the interest scores of the selected comments.

Abstract: Systems and methods for classifying electronic information or documents into a number of classes and subclasses are provided through an active learning algorithm. In certain embodiments, seed sets may be eliminated by merging relevance feedback and machine learning phases. Such document classification systems are easily scalable for large document collections, require less manpower and can be employed on a single computer, thus requiring fewer resources. Furthermore, the classification systems and methods described can be used for any pattern recognition or classification effort in a wide variety of fields, including electronic discovery in legal proceedings.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for . In an aspect, a system ranks web resources and native applications based on web resource scores and normalized native application scores that are normalized to the web resource scores. The ranking is indicative of the relevance of each web resource and native application for a search operation relative to each other web resource and native application.

Abstract: Various embodiments relate to systems, methods, apparatus, and computer readable media for querying data providers in point of interest searches. In one particular embodiment, a method is provided that includes identifying a plurality of data sources each having a data source priority, the plurality of data sources comprising at least two of: a first reference source, a first structured knowledge base, and a first individual website. The plurality of data sources may then be queried for metadata associated with the POI search request, wherein an order of the querying is based on the data source priority for each of the plurality of data sources; and at least one source quality measured for at least one of the plurality of data sources. In further embodiments, a query order may be updated based on the measured source quality.

Abstract: Systems and methods are disclosed for identifying a set of social look-alike users from a plurality of users. In an embodiment, a first set of users is selected from the plurality of users based, at least in part, on one or more characteristics associated with the plurality of users. A degree of similarity is determined between the first set of users and the plurality of users. The plurality of users is ranked based on the degree of similarity and thereafter the set of social look-alike users is determined based on the ranking.

Abstract: A method of determining a similarity between records in a data set is provided. Data organized into a plurality of records is received. First characters associated with a field and a first record of the plurality of records are selected. The selected first characters are subdivided into a first sliding series of a defined number of characters. Second characters associated with the field and a second record of the plurality of records are selected. The selected second characters are subdivided into a second sliding series of the defined number of characters. A similarity score between the first sliding series and the second sliding series is calculated. Whether or not the first sliding series and the second sliding series are similar is determined based on the calculated similarity score.

Abstract: A scanner scans a group of documents. For example, the documents can be a group of invoices. The documents are received and processed. Objects (e.g., a text object, such as a word) and their locations are identified in each of the documents. Occurrences of similar objects in the identified locations between the documents are determined. A document sorting algorithm is applied to generate a score for each of the documents. The score for each of the documents is generated based on a number of occurrences of similar objects between the documents. The generated score of each of the documents is used to identify a template document. The template document is then used to cluster the documents.

Abstract: Techniques are disclosed for facilitating re-creation of an application collection of a source computing device at a destination computing device. The techniques include receiving a source application identifier indicative of a source application edition, the edition of the application being programmed for a source operating system. The techniques also include receiving an indicator of a destination operating system. The techniques further include determining a source canonical application corresponding to the source application edition based on the source application identifier, the source canonical application being a representative of one or more application editions including the source application edition.

Abstract: A method and system is disclosed for tagging a latent object with selected tag recommendations, including a set of content objects wherein each object is characterized by an associated set of content features. An annotation relationship is determined between the features and a pre-determined tag for the each object, the relationship being defined by a graph construction representative of an affinity relationship between each pre-selected tag and content object to a selected query. A plurality of the annotation relationships are ranked based upon a relevance of the preselected tags to the content features in response to a new query for assigning a new tag to the each object, so that a suggested tag is made from the ranking whereby the suggested tag is determined as a most likely tag for annotating the content object.

Abstract: Methods and systems for arranging and searching a database of media content recordings are provided. In one example, a method is provided that comprises receiving a sample of media content, and performing, by a computing device, a content recognition of the sample of media content using a data file including a concatenation of representations for each of a plurality of media content recordings. In other examples, another method is provided that comprises receiving media content recordings, determining a representation for each media content recording, concatenating by a computing device the representation for each media content recording as a data file, and storing by the computing device a mapping between an identifier for a respective media content recording and a global position in the data file that corresponds to the representation of the respective media content recording.

Abstract: Example apparatus and methods concern dynamically expiring crowd-sourced content (CSC) in a crowd-sourced database. An example apparatus may include logic for acquiring the CSC, where the CSC is data produced by a mobile device concerning a point of interest. The example apparatus also includes logic for producing an evaluation of the CSC and logic for determining an expiration criteria based on the CSC, the evaluation, and the user. The CSC may be data about a point of interest. The evaluation may be based on the completeness, timeliness, or contents of the CSC. The expiration criteria may be established based on the evaluation of the CSC and a user profile. The expiration criteria or user profile may be manipulated based on confirmation or repudiation of the CSC by a different user or by curation of the CSC.

Abstract: A structured collection of message elements comprising message elements and oriented child-parent links each connecting a message element to a parent message element is provided. Each message element comprises a message content and metadata including an author identity and a timestamp. The message contents are parsed to generate appreciative phrase marks assigned to the message elements. An appreciative phrase mark is generated in response to detecting that the parsed message content of a later message element comprises a string of characters that matches an entry within a predefined dictionary of regard-expressing phrases. The appreciative phrase mark is assigned to an earlier message element that is connected to the later message element by a sequence of child-parent links. The metadata is parsed to detect the marks and further regard indicators assigned to the message elements. Relevance scores of the message elements are computed as a function of the regard indicators.

Abstract: The present invention relates generally to a method and apparatus for searching for recommended music using the emotional information of music and, more particularly, to a method and apparatus that enable recommended music to be searched for using mixed emotions by extracting emotional values including a valence value and an arousal value from an input search condition when a predetermined search condition is input by a user, extracting an emotion rank combination corresponding to the extracted emotional value information using an emotion model that includes mixed emotions corresponding to the emotional values, searching a music emotion DB for music information corresponding to the emotion rank combination, and outputting a recommended music list based on the results of the search, thus improving the user's satisfaction with the results of the search.

Abstract: Computer-implemented methods and computer systems for automatically managing stored checkpoint data are described. The method includes accessing a first user defined time period. The first user defined time period is related to a plurality of stored checkpoint data, and each checkpoint data of the plurality of stored checkpoint data has an associated storage time. Further, the method includes identifying a first set of checkpoint data having storage times that are within the first user defined time period. Moreover, the method includes identifying a second set of checkpoint data having storage times that are older than the first user defined time period. In addition, the method includes pruning the second set of checkpoint data according to a user specified process in proportion to storage time of each checkpoint data of the second set of checkpoint data. The older stored checkpoint data is more heavily pruned over recent stored checkpoint data.