Abstract: A “Facet Recommender” creates conversational recommendations for facets of particular conversational topics, and optionally for things associated with those facets, from consumer reviews or other social media content. The Facet Recommender applies a machine-learned facet model and optional sentiment-model, to identify facets associated with spans or segments of the content and to determine neutral, positive, or negative consumer sentiment associated with those facets and, optionally, things associated with those facets. These facets are selected by the facet model from a list or set of manually defined or machine-learned facets for particular conversational topic types. The Facet Recommender then generates new conversational utterances (i.e., short neutral, positive or negative suggestions) about particular facets based on the sentiments associated with those facets. In various implementations, utterances are fit to one or more predefined conversational frameworks.

Abstract: A method and a system for selecting a potentially erroneously ranked document in a set of search results responsive to a query comprising receiving the set of search results from the search engine server, each document of the set of search results having a relevance score and a feature vector generated by an MLA, computing for each possible pair of documents a first parameter indicative of a level of difference in the relevance scores of the documents of the pair of documents and a second parameter indicative of a level of difference in the feature vectors of the documents of the pair of documents, computing a verification score based on first parameter and the second parameter, the verification score indicative of a level of misalignment between the relevance scores and the feature vectors, selecting and marking the pair of documents associated with an extreme verification score for verification.

Abstract: A managing device (10) includes an editor (11) that causes a store to edit a product page for the store to sell a product, and to allocate a category name to the product page, a presenter (12) that searches a result page matching a query entered by a customer through the edited product page, and presents the result page to the customer, a register (13) that registers, when the product is purchased by the customer through the presented result page, a character string contained in the query with a page history in association with the result page, and a calculator (14) that calculates, based on a registration frequency with the page history for the category name allocated to the result page in association with the page history, a parameter indicating whether or not the allocation of the category name to the result page is suitable.

Abstract: Methods, computer program products, and systems are presented. The method computer program products, and systems can include, for instance: obtaining textual based message data of a messaging system; processing the textual based message data to determine one or more sentiment parameter associated to the textual based message data; and outputting an alert based on a result of the processing.

Abstract: A method may include identifying a shared usage of a first network address and a second network address in a predetermined period of time. The method may also include clustering the first network address and the second network address based on the shared usage. The method may include determining a weighting factor between the first network address and the second network address based on the shared usage. The method may further include receiving a request that includes the second network address from a client device. The method may include determining that the request for the electronic activity does not include suspicious activity based on the first network address and the second network address being in the cluster. The method may further include permitting the electronic activity based on the determination that the request for the electronic activity does not include suspicious activity.

Abstract: A solution is provided for detecting in real time video highlights in a sports video at a mobile computing device. A highlight detection module of the mobile computing device extracts visual features from each video frame of the sports video using a trained feature model and detects a highlight in the video frame based on the extracted visual features of the video frame using a trained detection model. The feature model and detection model are trained with a convolutional neural network on a large corpus of videos to generate category level and pair-wise frame feature vectors. Based on the detection, the highlight detection module generates a highlight score for each video frame of the sports video and presents the highlight scores to users of the computing device. The feature model and detection model are dynamically updated based on the real time highlight detection data collected by the mobile computing device.

Abstract: Described herein are a system and methods for providing relevant search results to a user using an automated user assistant to update filters used as search parameters. In some embodiments, an automated user assistant may initiate a conversation with a user detected interacting with an item. The user may provide feedback that indicates one or more attributes relevant to the user's interests with respect to the currently viewed item. In some embodiments, the user may also provide an indication of how values associated with those attributes would be made more relevant to the user. Various filters associated with the attributes may be updated based on the received user feedback. The updated filters may then be provided as search parameters to a search engine. A set of search results returned by the search engine may be provided to the user.

Abstract: An ontology is automatically generated for a set of data from multiple data sources. A semantic network of known concepts serves as an ontology template for a target domain of knowledge with known concepts defined as base entity types. Logical groupings of data and associated technical metadata are read from the data sources. Data discovery techniques are applied to detect semantic and/or syntactic classification of data attributes in the logical groupings of data. For each of the logical groupings of data, an entity type with properties for the data attributes are generated and the generated entity type is added to the semantic network with classifications of the properties as derived from the applied data discovery techniques. Semantic meanings of the generated entity types are generated and associated with other entity types within the semantic network and the semantic network is output as a resultant ontology for the set of data.

Abstract: Systems, methods, and non-transitory computer-readable media can obtain a plurality of content items that are available to be presented in a content feed of a first user, wherein each content item has a corresponding lifespan that indicates when the content item expires and is no longer available for presentation. A content feed having at least a first content item from the plurality of content items can be provided to a computing device of the first user. A determination can be made that at least one second user has provided feedback for the first content item, the feedback indicating an up-vote or a down-vote of the first content item. The lifespan of the first content item can be adjusted based at least in part on the feedback.

Abstract: Information extraction methods for use in extracting values from unstructured documents for predetermined or user-specified attributes into structured databases are provided herein. Methods include (a) automatically training machine learning models for extracting values from unstructured documents such that the values of the attributes are known for those training documents but the locations of the values in the documents are not known, (b) making a sustained connection between structured databases and unstructured documents so that the data across those two types of data stores can be cross-referred by the users any time, (c) a graphical interface specialized for rich user feedback to rapidly adapt and improve the machine learning models.

Abstract: A computing device analyzes the transaction entries in a transaction log to identify related commands associated with performing a set of data operations. Commands are considered to be potentially related if the commands are executed within a predetermined timespan. Sets of potentially related commands are then grouped together into corresponding candidate patterns, and further analyzed in view of additional information to determine a probability that the potentially related commands of a candidate pattern are actually related. A confidence value indicating that probability is also determined. Application management tasks, such as database optimization and recovery tasks, for example, may then be performed based on the candidate patterns that meet or exceed a predetermined threshold.

Abstract: Systems and methods are provided for generating and using classification models to automatically classify pages or other source content as including text about one or more real-world events. Generating the classification models may include analyzing text content of a large number of different pages from both a reference source and from more dynamic sources, such as from publisher sources via a network. Features for training classifiers may be determined based in part on the top n-grams identified among pages that have been determined to be associated with a given event.

Abstract: A method is provided that includes receiving a plurality of messaging signals for a user. The method further determines an elapsed time from the sending of a previous message to the user. The method also includes determining a priority order for the plurality of messaging signals, examining the plurality of messaging signals according to the priority order to identify a valid messaging signal, and sending a message to the user based on the valid messaging signal.

Abstract: In one embodiment, a network device (e.g., a RPL router) executes fast local RPL recovery in a low power and lossy network (LLN). The network device, in response to becoming an orphan in a directed acyclic graph (DAG) topology, can utilize the data plane to maintain at least some data traffic by randomly forwarding the data traffic to identified neighbor devices, while eliminating children from the list of forwarders and by finding successors that can be used for re-parenting. Hence, when a RPL network device having lost its last feasible parent can avoid data loss and accelerate a re-parenting process using local repair in the data plane instead of the control plane of the routing protocol used to establish the DAG topology.

Abstract: A feature term classification method executed by a processor included in an information processing apparatus including a display device and a memory, the feature term classification method includes extracting a feature term from a Web page displayed on a screen of the display device; executing image searching using the extracted feature term; classifying the feature term based on whether an image has been detected as information corresponding to the feature term in the image searching; storing the information corresponding to the feature term in the memory; when a new feature term has been input, extracting a Web page corresponding to the new feature term from the stored information corresponding to the feature terms; and displaying the extracted Web page.

Abstract: Some embodiments relate to assigning individual log messages to clusters. An initial cluster assignment may be performed by applying a hash function to one or more non-variable components of the message to generate an initial cluster identifier. Subsequently, clustering may be further refined (e.g., by determining whether to merge clusters based on similarity values). An interface can present a representative message of each cluster and indicate which portions of the message correspond to a variable component. Particular inputs detected at the input corresponding to one of these components can cause other values for the component to be presented. For a given cluster, timestamps of assigned messages can be used to generate a time series, which can facilitate grouping of clusters (with similar or complementary shapes) and/or triggering alerts (with a condition corresponding to a temporal aspect).

Abstract: A server, method, and non-transitory computer readable medium for classifying a query into one of a plurality of classes are provided. The server includes a network interface, a memory storage unit and a processor. The method involves receiving a query applying a plurality of support vector machine models, calculating a probability, and determining a top class. The non-transitory computer readable medium is encoded with codes to direct a processor to carry out the method.

Abstract: A system and method for building, maintaining and verifying an ontology are disclosed. The ontology may be build based on a meronymy and a taxonomy. In one example, the generated ontology may be an automotive mechanical systems ontology and the meronymy is a vehicle systems meronymy and the taxonomy is a universal parts taxonomy. The system may utilize consistent lexical derivation of the elements in the generated taxonomy.

Abstract: A task managing device facilitates and assists in execution of tasks relating to an accounting process. The device refers to management information in which task items to be completed are defined. The task items are displayed on a display screen in association with checkboxes which are selectable by a user to input completion of the respective task items. If a checkbox is selected to input completion of the associated task item, the device determines, based on the management information, whether an accounting process task needs to be performed for the completed task item. If the accounting process task needs to be performed, the device switches the display screen to an accounting process screen, and enters past data associated with a predetermined data input item in a past accounting process task in data input regions corresponding to the accounting process task.

Abstract: The present invention is directed towards systems and methods for providing a microdocument framework. The method and system includes receiving a plurality microdocuments and detecting content data for each of the plurality of microdocuments. The method and system further includes indexing at least a portion of the plurality of microdocuments based on the detected content and performing a searching operation using the content data associated with the microdocument data to determine a microdocument set. Thereupon, the method and system performs at least processing one operation on the microdocument set.

Abstract: In the case of a platform device (100) located at a web-site and capable of forming a network with a plurality of ID-detectable users or participants for gathering and processing items of information stored in memory spaces (110) of a platform (100) which are in each case code-identifiable and assigned to a plurality of different topics, there takes place an automated distribution of items of information collected by the users or participants in that the storage spaces (110) on the platform (100) are in each case formed by a dual unit (DuU), a dual unit (DuU) comprising in each case a first storage space (111) assigned to a predefined topic and provided with initial information (114) formulated by an initial participant but not editable by any other individual user or participant, together with a second storage space (112) that is assigned to the first storage space (111) and is editable by any one individual user or participant and is construed for the importation, by the plurality of users or participants,

Abstract: An active learning framework is operative to identify informative questions that should be added to existing question-answer (Q&A) pairs that comprise a training dataset for a learning model. In this approach, the question-answer pairs (to be labeled as “true” or “false”) are automatically selected from a larger pool of unlabeled data. A spatial-directed clustering algorithm partitions the relevant question-answer space of unlabeled data. A margin-induced loss function is then used to rank a question. For each question selected, a label is then obtained, preferably by assigning a prediction for each associated question-answer pair using a current model that has been trained on labeled question-answer pairs. After the questions are labeled, an additional re-sampling is performed to assure high quality of the training data. Preferably, and with respect to a particular question, this additional re-sampling is based on a distance measure between correct and incorrect answers.

Abstract: In some embodiments, apparatus, systems and methods are disclosed that utilize label data to assist in performing a retail store function. For example, label data may be utilized to provide an image of a label to help a store associate perform a bin, pick, stock or price change request. In some forms, the label data comprises a print data file which may be used to display the image of the label to the store associate. In other forms, the label data may comprise a larger data set representing the label image itself such as bit map data. In yet other forms, a hybrid of both image data and print file data may be used. In addition, timestamp and location data may also be used to help assist in performing the requested store function more quickly and efficiently.

Abstract: Techniques and mechanisms are disclosed to increase the availability of summary data within a clustered data intake and query system by replicating the summary data within the cluster. In general, summary data may store “pre-computed” results for one or more search queries and can be used by indexers of a cluster to process subsequent instances of the same search queries. At a high level, replication of summary data within a cluster may include ensuring that each instance of summary data created by an indexer of a cluster is replicated to other indexers within the cluster that store copies of the same grouped subset(s) of data to which the summary data relates. In this manner, if one or more indexers of an indexer cluster fail, other indexers of the cluster can make immediate use of replicated copies of the summary data without re-creating it.

Abstract: Given a search result, the set of authors-identifiers appearing in it can be determined by use of a hash function, and an array-type data structure called an audience fingerprint (AF). The AF has as many storage locations as the hash function has possible output values. The number of possible output values is chosen to be large enough, with respect to the maximum number of unique authors expected in any one search result, to create a very high probability of a unique output value for each unique author-identifier that is hashed. At the AF location, addressed with a hash value, is stored an indicator that the author-identifier is present. The indicator can be a single bit, simplifying set operations on AFs. When not in working memory, an AF can be stored as a compacted sparse array. The actual author-identifiers present can be determined, from an AF, with an inverse hash function.

Abstract: A digital downloading jukebox system including a mechanism for delivering custom services to a recognized user is provided. For example, information specific to a recognized user may be stored and optionally may include a recognized user avatar representative of the recognized user. The user avatar may be an image, video, and/or animation, which may be displayed on and/or played through the jukebox. The user avatar may be associated with transactions associated with the user. For example, an avatar may be displayed when a playlist of the recognized user is played, when a message is sent, etc. In other examples, the avatar may introduce instances of media by playing an audio and/or video message, and the avatar may sing, dance, etc. while an instance of media is playing.

Abstract: A computer-implemented method for managing portfolio-relevant content is implemented by a content management computer system coupled to a memory. The method includes receiving a set of assets associated with an investor account, transmitting an investment content request for an investment content associated with at least one asset of the set of assets, receiving the investment content including a plurality of investment content items, determining a plurality of content indicators associated with each investment content item, determining a set of relevant investment content including a plurality of display parameters based at least partially on at least a portion of the investment content and at least a portion of the set of assets, and providing the set of relevant investment content for display on a user device based on at least a portion of the display parameters.

Abstract: In the case of a platform device (100) located at a web site and capable of forming a network with a plurality of ID-detectable users or participants for the purpose of gathering and processing items of information stored in code-identifiable memory spaces (110) of a platform (100) assigned to a plurality of different topics by the users or participants, a passive automated distribution of information collected by the users or participants is made possible in that the memory spaces (110) on the platform (100) are each formed by a dual unit (DuU), which dual unit (DuU) comprises a first memory space (111) that is assigned to a predefinable topic and provided with initial items of information (114) which has been formulated by an initial participant but which are not editable by any other individual user or participant and a second memory space (112) assigned to the first memory space (111), which second memory space (112) is editable by any individual user or participant and is construed for the inclusion, by

Abstract: An automatic proximity discovery area technique uses available social network information to establish a proximity service area and to select and enable one or more host user equipment devices for device-to-device proximity services. The automatic proximity discovery area technique reduces power consumption for device-to-device proximity service discovery and communications. The technique uses social network messages to detect events and initiate discovery of the device-to-device communications in a proximity discovery area without constant querying or permanent discovery facilities, thereby reducing power consumption of device-to-device proximity services.

Abstract: Disclosed is a method and system to provide the trending news stories to the plurality of groups based on the plurality of group members' existing conversations. Group admin registers the group by adding the phone number or email address. The impersonation access acquired from the group admin to read through the existing conversation by the group members. The present invention retrieves the top words from the existing group conversations among the group members and retrieves the trending news stories for the individual words to format them and presented it to the group. The group members' interactions on the presented trending news stories are being monitored and updated it to the respective social media accordingly.

Abstract: Aspects of the disclosure are directed to DC power management. A sequencer may be configured to execute a first command, wherein the first command is associated with a unique group tag; compare the unique group tag to a master group tag; determine if an interrupt is detected; lock the master group tag to yield a locked master group tag; execute a second command, wherein the second command is associated with the locked master group tag; determine that an end of commands in the locked master group tag is reached and execute a sequence jump through command to put a processor back to a regular power state.

Abstract: Multiple social network postings that originate from a single social network user are detected within a social network. The multiple social network postings that originate from the single social network user are analyzed using configured posting relationship identification criteria. Based upon the analysis of the multiple social network postings that originate from the single social network user using the configured posting relationship identification criteria, a determination is made that at least two of the multiple social network postings that originate from the single social network user include related content. The content of the at least two social network postings determined to include the related content is concatenated into a single consolidated social network post.

Abstract: A population formation method is disclosed. Keywords are extracted from public information of providers included as elements in a first provider group. Each element is calculated based on a predetermined attribute value. A first attribute is for the providers of the public information. The attribute value is changed with time. Each of rules set for duplicate keywords is to determine one of the attributes by using one of the duplicate keywords. Provider groups are formed for new public information based on the duplicate keywords and the rules. A provider group having a similar relationship with a first provider group is specified by a distribution of the attribute value of a different attribute from the first attribute. A new provider group corresponding to the first provider group is formed by the providers, for whom the attribute value of the first attribute corresponds to the predetermined attribute value.

Abstract: Embodiments of the present invention provide techniques for analyzing technical documents against known art. One or more machine learning classifiers are trained by a training set. References are ingested and assembled into a corpus. An input query is processed by natural language processing, and the classifiers are used to retrieve relevant documents and references, providing solutions to the input query. Anti-patterns are identified and provided to a user in the form of a report to alert the user to potential problems with the input query.

Abstract: Methods, apparatus, systems and articles of manufacture to perform identity matching across audience measurement systems are disclosed.

Abstract: According to one embodiment of the present invention, a method for name searching in mixed-entity lists is provided which comprises dividing a mixed list of entities into a plurality of entity-specific lists. A name to be searched is then categorized into a category and a specialized search logic is applied to the name to be searched. The specialized search logic is selected to be adapted to the category and uses a one of the entity-specific lists that corresponds to the category of the name to be searched. A shared search logic may also be employed, which is used for all names to be searched.

Abstract: The present invention suppresses feature amounts of faces to be registered in a collation database, thereby preventing the collation database from being bloated. An information processing apparatus of the present invention includes an acquisition unit configured to acquire a video, a detection unit configured to detect at least one face of the same person from a plurality of frames of the acquired video, a classification unit configured to classify the detected faces into a plurality of predetermined groups, a selection unit configured to select, from the faces classified into the groups, not more than a first predetermined number of faces for each group, wherein the first predetermined number is an integer of not less than 2, and a registration unit configured to register feature amounts of the selected faces in a database.

Abstract: A document such as a book or textbook includes multiple sections such as chapters. Concept phrases are determined for each of the sections based on the text of each section. A set of content items such as videos is received, and each content item is associated with one or more queries that were submitted by users who were provided the content item in a set of search results. These queries are processed to determine concept phrases that are associated with the content items. The content items and their associated concept phrases are compared with the concept phrases associated with the sections to determine, for some or all of the content items, a minimum subset of the sections whose associated concept phrases cover most of the concept phrases that are associated with the content item. The content items are inserted or linked with the sections in their corresponding minimum subsets.

Abstract: Project asset and preference sharing techniques are described. In one or more embodiments, a request is received to assign a project asset or preference to a member of a team. The received request includes an identifier of a project for which the project asset or preference will be assigned. To assign the asset or preference specified in the request, a list of teams to which the member belongs is ascertained. The assets and preferences associated with each of the teams are then checked for the identifier to identify the project assets and preferences associated with the project. Once the project assets and preferences associated with the project are identified, a response is generated for communication to the member. The response is configured to include indications of the identified project assets and preferences that enable the member to access the identified project assets and preferences via the application, such that the member is also given access to the assigned project asset or preference.

Abstract: Systems and methods of the present disclosure provide technology to identify when network-connected devices are likely infected with malware. Network communications are be monitored during a specific time window and a graph is created for a conditional random field (CRF) model. Vertices of the graph represent devices connected to the network and an edge between two vertices indicates that one or more network communications occurred between two devices represented by the two vertices during the time window. Network devices can report observations about network behavior during the time window and the observations can be used as input for the CRF model. The CRF model can then be used to determine infection-status values for the network devices.

Abstract: Provided are techniques for providing and using pivot facets for text mining and search. A metadata field is mapped to a pivot facet. In response to executing a first query, at least a first document and the pivot facet are displayed. In response to selection of the pivot facet, a second query that includes a value of the metadata field is generated, and the second query is executed to locate at least a second document that includes the value of the metadata field.

Abstract: Resources are required to satisfy various needs and wants of people, businesses, and machines. Resources come in the forms of time, talents, money, materials, energy, services, people, knowledge, communication, and other tangible and intangible assets. When both the capacities and the needs of multiple resources are stored in a way that allows for them to be connected together using computers, they can be efficiently and effectively matched. This matching creates shared value, which has potential academic, economic, societal and philanthropic benefits. Connected computer system(s) can query and match resources together in a way that is mutually beneficial. While a common lexicon is the simplest way to perform the matching, natural language processing, machine translation, or use of similar technologies may be optimal. Any method of collecting these inputs should be able to handle one or multiple capacities, and one or multiple needs.

Abstract: A device common model interface is described that translates a schema describing a resource in a fabric network into a common object model and one or more target-specific data models, using a specification of a target application environment and a specification of a communication protocol. Elements of the schema that are useful for an application are extracted to generate the common object model and the one or more target-specific data models are provided to the application. The application interacts with an interface of the resource using the API of the target-specific data model, which performs operations that are specific to communicating with the resource in the target application environment and using the communication protocol.

Abstract: In one embodiment, a method is performed by a computer system. The method includes receiving a request from a user to access particular content. The method further includes determining a trust measure of the user, wherein the trust measure is based, at least in part, on an analysis of logged user-initiated communication events of the user on a plurality of communications platforms. In addition, the method includes accessing a self-service access policy applicable to the particular content. Further, the method includes ascertaining, from the self-service access policy, a trust threshold applicable to the particular content. Moreover, the method includes, responsive to a determination that the trust measure fails to satisfy the trust threshold, automatically denying access by the user to the particular content.

Abstract: In one embodiment, one or more computing devices determine, substantially in real time and locally at a mobile device, one or more current dynamic operating conditions of the mobile device related to energy consumption by the mobile device. The computing devices identify a pre-determined device state that corresponds to the current dynamic operating conditions of the mobile device and determine, based at least in part on the pre-determined device state, an action for the mobile device to take, the action being related to energy consumption by the mobile device.

Abstract: A method is provided for receiving and rendering web content on a client device having an installed actor app. Web content containing an embedded application tag is requested from a server. The web content is received from the server on the client device. The actor app receives and injects a value in the embedded application tag from dynamic data received with the web content. The web content is displayed together with the value of the dynamic data on the client device.

Abstract: Disclosed are various embodiments for identifying correlations between child nodes of a taxonomy. A correlation service identifies a correlation between parent nodes of the taxonomy. A data set such as an interaction history is filtered according to the parent node correlation. The filtered data set is then used to identify correlations between the respective child nodes of the correlated parent nodes.

Abstract: In one aspect, a method includes receiving an indication that a first user has interacted with a second user in a social network, calculating a social affinity score between the first user and the second user based on first user information, second user information, or relationship information between the first user and the second user and scheduling delivery of a notification to the recipient based on the social affinity score. Systems and machine-readable media are also provided.

Abstract: A video transmission method is disclosed herein. A target video segmentation is transmitted to a piece of user equipment relatively quickly, so that caching efficiency of a caching device can be improved, and a processing burden of the caching device can be reduced. The method includes: receiving, by a gateway device, a target URL sent by a second user equipment, and determining, from a caching device, a second video corresponding to the target URL; determining a segmentation part of the target URL according to the segmentation part extraction model; and determining a target video segmentation from the second video according to the segmentation part of the target URL, and sending the target video segmentation to the second user equipment.

Abstract: Data mining systems and methods are disclosed for item recommendation based on frequent attribute-values associated with items. The system may determine commonalities in item attribute-values based on user transactions and identify frequent attribute-value tuples that include attribute-values that frequently co-occur in user transactions. The system may associate user interests with the frequent attribute-value tuples and recommend items to target users based thereon.