Abstract: Methods, systems, and machine-readable media for identifying synthetic media file examples to train a supervised machine learned ranking algorithm to rank relevance of media files to a search query are provided. In one aspect, a method includes identifying a search query for a search engine for a collection of media files from previously received search queries, and selecting at least one training media file from the collection as a synthetic negative example for the search query. The method also includes providing a training set to a supervised machine learned ranking algorithm. The training set includes an identification of the search query, a copy of the training media file, and a first indicator that the training media file is a synthetic negative example for the search query. The method further includes providing, to the algorithm, the search query and the collection, and receiving, from the algorithm, a ranking of the collection.

Abstract: Described is a system and method for enabling visual search for information. With each selection of an object included in an image, additional images that include visually similar objects are determined and presented to the user.

Abstract: Unstructured data are processed in response to a request for information to derive attributes. The attributes include a subset of variables utilized to determine the requested information. The attributes are ranked based on a set of relevance factors for determining the requested information. One or more attributes are selected for presentation to a user based on the ranking to receive modifications and produce modified attributes. The requested information is determined based on the set of variables including the modified attributes.

Abstract: An approach for filtering data is presented. A relationship between first and second entity-metadata elements specifying a person and a vehicle, respectively, and between the person and the vehicle is determined. Representations of the first and second entity-metadata elements are displayed within a regular polygon that includes locations indicated by a geospatial tag that includes location information about the person extracted from profile information describing the person and by other geospatial tags included in metadata obtained from data extracted from streaming data and data at rest. The metadata includes contextual information that specifies an activity included in a domain of knowledge associated with law enforcement. Based on hidden Markov and support vector machine models, a frequent pattern growth algorithm, and a Kohonen map, another activity of the person is predicted.

Abstract: In a method for evaluating a performance of a recommender system, a session of M content items consumed by a user can be identified. The session can be defined such that, for each of a first (M-1) content items in the session, a delay time between a completion of consumption of one of the first (M-1) content items and a commencement of consumption of a directly next content item in the session is less than a predetermined duration of time. Content items in a content item-base can be ranked, based on the first (M-1) content items, according to a likelihood of each content item in the content item-base to be consumed next. A session metric can be generated for each content item in the content item-base. The recommender system can be trained based on the session metric for the Mth content item.

Abstract: Systems and methods are provided for operating a user-controllable adventurousness parameter. An item pool is accessed, where the item pool contains a plurality of items, and where each item is associated with a parameter. A baseline position is identified, where the baseline position is associated with a particular value of the parameter. A user-controllable adventurousness parameter is received, where the adventurousness parameter indicates a user preference for a next item relative to the baseline position. The next item is selected based on the baseline position and the adventurousness parameter, and the next item is provided via a computer network.

Abstract: Methods and systems for extracting intents and intent profiles of users, as inferred from the different activities they execute and data they share on social media sites, and then (i) monetization of such intents via targeted advertisements, and (ii) enhancement of user experience via organization of their contact lists and conversations and posts based on their content and conceptual context.

Abstract: A system and method that identifies a search term used in search queries of a collaboration platform is disclosed. The system and method obtain historical data associated with the search term. The historical data includes first data indicative of a number of times a content item has been returned in search results responsive to the search queries using the search term. The historical data includes second data indicative of a number of times the content item was selected from the search results and interacted with by users for a threshold amount of time. The system and method generate a score for the search term based on the historical data, wherein the score is associated with the search term and the content item.

Abstract: A computer implemented method for estimating quality of document retrieval comprising: retrieving from a corpus of documents stored on at least one storage a plurality of digital documents which comply with a document retrieval query according to a retrieval model; computing a plurality of retrieval scores each calculated for one of the plurality of digital documents using a relevance function scoring a relevance of one of the retrieved plurality of digital documents to the query; computing a calibrated weighted product model (WPM) estimator by calculating a combination of the plurality of retrieval scores weighted according to a plurality of retrieval features of the corpus and/or the query and/or a document, wherein the plurality of retrieval features are weighted according to a relative importance; and using the calibrated WPM estimator to score the plurality of digital documents' relevance to the query.

Abstract: A method and system for true product word recognition using a computing device. The method includes: obtaining title words from titles of products; determining word vectors based on a word matrix, where the word matrix is constructed based on a number of cooccurrence of each of the title words and all the other title words; determining title vectors based on a title matrix, where the title matrix is constructed based on the number of cooccurrence of each of the title words and all the other title words, and the title words in each of the titles; calculating scores for the title words in corresponding one of the titles based on the corresponding word vector and title vector; and recognizing true product words of the corresponding one of the titles based on the scores.

Abstract: A system and method are described herein for improving content-based (CB) filtering to find other items similar to a particular item. For example, the system and method can operate to (1) collect ground truth from users, experts or non-experts for a subset of items in a library; (2) build similarity model(s) against the collected ground truth and evaluate this/these similarity model(s) for generalizability to all items in the library, and (3) build recommendations for users based on knowledge elicited from the users themselves, combined with the similarity scores in the built similarity model(s).

Abstract: Implementations include systems and methods for querying a data graph. An example method includes receiving a machine learning module trained to produce a model with multiple features for a query, each feature representing a path in a data graph. The method also includes receiving a search query that includes a first search term, mapping the search query to the query, and mapping the first search term to a first entity in the data graph. The method may also include identifying a second entity in the data graph using the first entity and at least one of the multiple weighted features, and providing information relating to the second entity in a response to the search query. Some implementations may also include training the machine learning module by, for example, generating positive and negative training examples from an answer to a query.

Abstract: Technologies are described herein for the identification of documents based on location, usage patterns, and content. In some configurations, techniques disclosed herein cause documents to be identified that are relevant to the location of the user and the current context of the user. Some illustrative configurations involve identifying documents that are associated with a particular location. In addition to using location information, other data can be analyzed to identify documents that are relevant to the current location of the user and/or the current context of the user. The other data can also include data such as, but not limited to, calendar data, document data (e.g., contents of documents, metadata associated with documents), organizational charts, and contact lists. The documents that are relevant to the meeting participants and the subject of the meeting can then be presented to the user for easy access.

Abstract: An information processing device for obtaining content is provided. The information processing device includes circuitry configured to receive audio/video (AV) content via a broadcast signal. The circuitry is configured to receive from a provider of the AV content, provider-side preference, demographic or interest (PDI) information. The provider-side PDI information is associated with further content. The circuitry is configured to obtain user-side PDI information, and determine whether the provider-side PDI information and the user-side PDI information match. The circuitry is configured to obtain the further content when the provider-side PDI information and the user-side PDI information are determined to match, including calling at least one application programming interface (API).

Abstract: A linguistic profiling method, system, and computer program product include capturing a language input of an individual user, analyzing the language input to detect an attribute of the language input, and synthesizing the attribute to characterize a feature of the individual user.

Abstract: Embodiments of the present systems and methods may provide a fundamentally different approach to process the ‘search key words’ from users. For example, the process may progress through a series of steps that will understand ‘WHY’ a user is looking for the information instead of just ‘WHAT’ the user is searching. This will give the present techniques a very different way to look for the information and generate the search outputs. For example, a method may provide search results to a user of a computing device, the method may comprise receiving at least one search term from the user via the computing device, collecting information relating to the user other than the at least one search term, and providing search results to the user via the computing device based on the search term and on the collected information relating to the user.

Abstract: A delivery coordination system selects restaurant listings for presentation to a user by filtering out restaurants that are unlikely to be of interest to the user, while ensuring that restaurant listings are selected to encourage the success of all restaurants using the delivery coordination system. In response to receiving the listings request from a client device, the delivery coordination system selects a filtered set of restaurant listings from the plurality of restaurant listings stored by the delivery coordination system by applying filtering criteria to the plurality of stored restaurant listings. The delivery coordination system generates conversion scores for the restaurant in the filtered set of restaurant listings and selects restaurant listings to present to the user from the filtered set of restaurant listings based on selection factors, which can include the generated conversion scores. The delivery coordination system transmits the selected restaurant listings to the client device.

Abstract: According to one aspect of the present disclosure, a method is provided for advertising on a content-screened web page. The method includes receiving an impression request for a URL for which an advertising impression is desired; screening HTML content of a web page identified by the URL; generating a rating for the URL based on the HTML content of the web page; assigning the rating to the impression request; and serving an advertising impression on the web page based on the rating.

Abstract: An online system generates dynamically optimized sponsored content for a target user of the online system. Each sponsored content item comprises optimal creatives selected for a target user from a set of creatives provided by a content provider. Each type of creative (e.g. title, image, video, descriptive text), has a trained creative model to generate a prediction score for a creative of the same type based on the features of the creative and the characteristics of the target user. The prediction score of a creative indicates the likelihood that the target user will interact with a sponsored content item that includes that particular creative. The online system selects a creative for each type and assembles the selected creatives into a sponsored content item for the target user. For a different user, a different sponsored content item having different creatives chosen from the same set of creatives is generated.

Abstract: Systems and methods describe presenting media to a user at a user device, determining a location relevant to the media, determining whether the user has visited a vicinity of the location by searching through information related to the user, and based on to determining that the user has visited the vicinity of the location searching through the user's personal collection of media in a time window corresponding to the user's visit to retrieve personal media related to the location, and generating for display, based on the searching, a list of personal media related to the location relevant to the media being presented at the user device.

Abstract: A method of assessing the legitimacy of a subject pharmacy website includes the steps of, from a network-enabled computing device, generating a prediction model comprising a list of pharmacy websites of known legitimacy and a first set of websites which contain at least one referring link to at least one of the list of pharmacy websites of known legitimacy, from the network-enabled computing device, collecting a second set of websites which contain at least one referring link to a subject pharmacy website whose legitimacy is unknown, isolating a subset of websites from the second set of websites based on the prediction model, comparing the subset of websites to the first set of websites linked to the list of pharmacy websites whose legitimacy is known, and determining the legitimacy of the subject pharmacy website based on the comparison to the first set of known legitimate pharmacy websites.

Abstract: A digital content acceleration system comprising: a keyed database for storing keyed data; a data retrieval engine that retrieves, in response to receiving an item of keyed data, one of i) search data indicative of a search history associated with the item of keyed data, ii) web history data indicative of one or more web pages accessed by a terminal device associated with the item of keyed data, or iii) both i) and ii); a parsing engine that extracts one or more attributes from the search data and the web history data; a terminal device network access engine that generates, based on the one or more attributes, a probability distribution for geographic locations; a content selection engine that retrieves, from a digital content provider, digital content associated with a particular geographic location of the geographic locations; and a digital content assembler engine that pre-assembles the digital content.

Abstract: In one embodiment, a method includes receiving a search query from a client system; modifying the search query by removing a first search term associated with a first object type from the search query; retrieving search results responsive to the modified search query from a first data store storing objects of the first object type; and sending one or more of the search results to the client system.

Abstract: User-directed identity verification over a network. In one embodiment, a method may include receiving identity information of a user, storing the identity information, calculating a confidence score for the identity information that indicates a level of confidence that the identity information belongs to the user, receiving a request for the identity information of the user, or for confirmation of possession of the identity information of the user, to be sent to an organization, sending an authorization request for the user to authorize the organization to access the identity information or to receive the confirmation of possession of the identity information, receiving a grant of authorization for the organization to access the identity information or to receive the confirmation of possession of the identity information, sending the identity information, or the confirmation of possession of the identity information, and the confidence score for the identity information.

Abstract: Machine-learning based multi-step engagement strategy generation and visualization is described. Rather than rely heavily on human involvement to create delivery strategies, the described learning-based engagement system generates multi-step engagement strategies by leveraging machine-learning models trained using data describing historical user interactions with content delivered in connection with historical campaigns. Initially, the learning-based engagement system obtains data describing an entry condition and an exit condition for a campaign. Based on the entry and exit condition, the learning-based engagement system utilizes the machine-learning models to generate a multi-step engagement strategy, which describes a sequence of content deliveries that are to be served to a particular client device user (or segment of client device users).

Abstract: A computer-implemented method according to one embodiment includes identifying a search query, determining a first set of categories associated with the search query, determining a second set of categories associated with preliminary results of the search query, and filtering the preliminary results of the search query based on an analysis utilizing the first set of categories and the second set of categories.

Abstract: Systems, methods, and non-transitory computer-readable media can determine a respective latent representation for each entity in a set of entities that are accessible through the social networking system, wherein a latent representation for an entity is determined based at least in part on a topic model associated with the entity, each latent representation for an entity having a lower dimensionality than a topic model of the entity. One or more candidate entities that are related to a first entity can be determined based at least in part on the respective latent representations for the candidate entities and the first entity. At least a first candidate entity from the one or more candidate entities can be provided as a recommendation to a user that formed a connection with the first entity.

Abstract: A computer receives a request for graphical display source code for a computerized graphical advertisement display, and retrieves seed animations including a plurality of seed animation features. The computer generates candidate animations based on the one or more seed animations, where the computer alters a first aspect of a seed animation to generate an altered seed animation having a plurality of altered seed animation features and the computer alters a second aspect of the altered seed animation to generate a candidate animation having a plurality of candidate animation features. The computer generates candidate animation scores based upon a context of the advertisement display and the plurality of candidate animation features. The computer selects an animation from the candidate animations based on the candidate animation scores and generates the graphical display source code based on the selected animation, a size of the advertisement display, and display capabilities of the user device.

Abstract: Systems and methods including one or more processing modules and one or more non-transitory storage modules storing computing instructions configured to run on the one or more processing modules and perform acts of receiving a plurality of user reviews of a product, performing topic modeling of the plurality of user reviews of the product to find a plurality of snippets within the plurality of user reviews each relating to at least one user attribute category of a plurality of user attribute categories, and facilitating a display on a device of a first snippet of the plurality of snippets proximate the product.

Abstract: Systems and techniques for graph-driven authoring in productivity applications can include using context of a user of a productivity application, the existing content of the user's document or an expression input with a command for finding relevant documents, and a date/time to obtain a set of results of relevant documents. The relevant documents include documents associated with people known to the user that contain content relevant to the topic(s) and/or entities from the context and that are current.

Abstract: A data management system and unit host applications and manage storage, finding and retrieval of typed items with support for tagging, connections, and situated queries. The data management system and unit have a content processor that receives content and processes the content to add one or more items into a particular cell, an item processor that receives one of an item to add to the cells in the storage unit and an item to update in the cells of the storage unit and a query processor that executes a query against the one or more cells in the storage unit.

Abstract: The technology described herein provides an efficient mechanism for quickly analyzing huge amounts of media content to find media content (hereafter “content” or “media content”) that is relevant to a user. The technology analyzes features of a curator to classify curators by interest and/or find curators with similar content recommendations. The curator data can be used to make curator recommendations to users based on the user's interests. The technology described herein collects curator data from multiple content sites and analyzes the data to identify curators that recommend similar content on different content sites.

Abstract: Systems and methods are described for presenting search results from multiple sources by grouping the results from some of the multiple sources, ranking each of the multiple sources and groups of sources, and not presenting duplicate results from lower ranked sources. In this way, the user is provided with search results that are distinct as opposed to presenting the same result multiple times when it is available from different sources.

Abstract: A profiling-based energy-aware recommendation method for a cloud platform at a cloud mediator interface may include: collecting usage data of a cloud node over predetermined intervals of time; generating and storing an energy usage profile for each node by using the collected usage data; acquiring real-time usage data from the cloud node; and comparing the real-time usage data with the generated energy usage profile so as to calculate a recommendation value.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for triggering rich results in response to queries. In one aspect, a method includes receiving a query. One or more search results are obtained from a first corpus. A rich result is triggered based on a score of the first-ranked search result if it meets a threshold relative to other search results. The rich result is populated with additional metadata about the first-ranked search result obtained from a second corpus. The rich result is provided in response to the query.

Abstract: Systems and methods for automatic user profile generation and authentication. One system includes a database storing user profile data for a plurality of users and an electronic processor communicatively coupled to the database. The electronic processor is configured to receive, from a computing device, a first usage indicator for a first user and a first user identifier for the first user, and to generate a probable user profile based on the first usage indicator and the first user identifier. The electronic processor is configured to retrieve, from the database, a plurality of previously-generated user profiles, and to determine, for each of the previously-generated user profiles, a confidence score based on the probable user profile. The electronic processor is configured to select an active user profile from the plurality of previously-generated user profiles when the confidence score exceeds a confidence threshold, and transmit, to the computing device, the active user profile.

Abstract: Systems and user interfaces enable integration of data items from disparate sources to generate optimized packages of data items. For example, the systems described herein can obtain data items from various sources, score the data items, and present, via an interactive user interface, options for packaging the data items based on the scores. The systems may include artificial intelligence algorithms for selecting optimal combinations of data items for packaging. Further, the interactive user interfaces may enable a user to efficiently add data items to, and remove data items from, the data packages. The system may interactively re-calculate and update scores associated with the package of data items as the user interacts with the data package via the user interface. The systems and user interfaces may thus, according to various embodiments, enable the user to optimize the packages of data items based on multiple factors quickly and efficiently.

Abstract: Systems, methods, and non-transitory computer readable media can identify a user associated with a device based on a subset of media content items on the device based at least in part on analysis of the subset of media content items. A relationship between the user and one or more other users depicted in the media content items can be determined. A recommendation relating to sending at least one media content item on the device to at least of the one or more other users can be generated based on the determined relationship.

Abstract: Techniques include: receiving a search query; identifying content responsive to the search query; identifying one or more members of a social networking service that have an association with identified content; and outputting, to a computing device, data corresponding to the one or more members, where data corresponding to the one or more members is for generating display areas, and where the display areas are for obtaining information about corresponding members and for display along with search results.

Abstract: A system and method for behavior influenced search ranking may include obtaining, via a network interface, a search term from a user device. An initial result including a first group of the user profiles may be generated based on user profiles from a social network in relation to the search term, the user profiles stored in a profile database. A rank of each of the first group of the user profiles may be determined based, at least in part, on interactions from an activity database corresponding to the first group of the user profiles, the activity database storing information indicative of activities related to the social network, the activities including the interactions. The user device may display a second group of the user profiles, including at least some of the first group of the user profiles, according to the rank of the first group of the user profiles.

Abstract: A method and system are provided for processing natural language user queries for commanding a user interface to perform functions. Individual user queries are classified in accordance with the types of functions and a plurality of user queries may be related to define a particular command. To assist with classification, a query type for each user query is determined where the query type is one of a functional query requesting a particular new command to perform a particular type of function, an entity query relating to an entity associated with the particular new command having the particular type of function and a clarification query responding to a clarification question posed to clarify a prior user query having the particular type of function. Functional queries may be processed using a plurality of natural language processing techniques and scores from each technique combined to determine which type of function is commanded.

Abstract: Techniques are provided for determining a user of a computing device who is likely to be unsatisfied with a value of a setting of a program running on the computing device. First information is collected about the value of the setting from a first population of users who invoke a routine to automatically reset the value of the setting to a default value. Second information is collected about the value of the setting from a second population of users who do not invoke the routine. Based on a comparison of the first information and the second information, a determination is made that a user in the second population is likely to be unsatisfied with the value of the setting.

Abstract: The present document describes a system for generating a graphical user interface for display by a client computer. The system comprises a server, with various components, for communicating with a plurality of client computers. The server is made to get in a database memory a plurality of pieces of conceptual information and a first threshold associated to a first user account, and a plurality of pieces of conceptual information and a second threshold associated to a second user account; to determine if the relative numbers of pieces of conceptual information that the first user account and the second user account have in common both exceed the first threshold and the second threshold, respectively; and to generate a graphical user interface dataset and sending the graphical user interface dataset to the first client computer for display, the graphical user interface suggesting a connection to the second user account.

Abstract: According to various embodiments of the present disclosure, an electronic device analyzes shared contents of all members in a circle created by a client user and generates feature information of the circle. The electronic device analyzes shared contents of each member in a buddy list created by the client user and generates feature information of each member in the buddy list. The electronic device calculates a similarity between the feature information of each member in the buddy list and the feature information of the circle, generates a similarity set, and generates a recommended buddy of the circle based on the similarity set. The electronic device prompts the recommended buddy of the circle to the client user.

Abstract: An online computer system is configured to receive requests from seekers for products or services and to provide the seekers and products a place to complete work once connected. The computer system searches through personal contacts of the seekers for trusted providers of the product or service request based on their proximity to the seekers. The computer system is configured to search through contacts within referral networks of the personal contacts for level-1 and level-2 providers of the product or service request. The level-2 providers might be more reliable because they are referred through the personal contacts of the seekers. The providers and referral contacts of the referrals can be rated and then provider candidates ranked according to the associated provider and referral ratings.

Abstract: Embodiments are directed towards managing changes to data. A modeling engine may provide a data model based on objects comprised of one or more versions of one or more properties. Each version of the properties may be associated with separate validity ranges over time during which each version is valid. A report for visualizing the data model at a point-in-time may be provided. One or more report values may be provided based on each version of the properties that have a validity range that includes the point-in-time. The report may be displayed to provide provides one or more visualizations based on the one or more report values.

Abstract: Embodiments of the present invention provide a system, method and computer program product for generating a succinct approximate representation of a time series. A method comprises determining at least one peak data value and at least one trough data value of an input series comprising a sequence of data values over time. The input series is partitioned into multiple segments comprising at least one rising segment that rises to a peak data value and at least one falling segment that falls to a trough data value. A sequence of segments that rise and fall alternately is generated based on the segments. A sequence of totals representing a succinct approximate representation of the input series is generated. Each total comprises a sum of data values for a corresponding segment of the sequence of segments.

Abstract: A method includes receiving, at a computing device, search results transmitted from a search system in communication with the computing device. The search results include a header for an application executable on the computing device and application access mechanisms associated with the header. Each application access mechanism has a reference to the application and indicates one or more performable operations for the application. The method further includes displaying, on a display in communication with the computing device, a graphical user interface including the header and user selectable access links grouped with the header. Each access link is associated with an application access mechanism for the application.

Abstract: Systems and methods are described herein for a media guidance application that facilitates presenting, to a user, media assets that match a search expression by generating for display media asset identifiers corresponding to media assets that are scheduled for transmission to a plurality of users during a specific time interval. The media asset identifiers corresponding to media assets that match the search expression are distinguished from media asset identifiers that correspond to media assets that do not match the search expression. Furthermore, media asset identifiers corresponding to media assets that match the search expression are distinguished from each other based on an attribute of each media asset that matches the search expression.

Abstract: A method and an apparatus for determining a morpheme importance analysis model is provided, which belongs to the field of computers. The method includes: acquiring at least two pieces of training data, each piece of training data including a query, any morpheme in the query, and an importance score of the any morpheme in the query; determining a feature value of each preset feature of each piece of training data; and determining a model parameter according to the feature value of each preset feature of all training data and importance scores of morphemes included in all training data, and determining a morpheme importance analysis model according to the determined model parameter.