Abstract: Disclosed in some examples are systems, methods, and machine readable mediums which allow for the automatic creation of a skills hierarchy. The skills hierarchy comprises an organization of a standardized list of skills into a hierarchy that describes category relationships between the skills in the hierarchy. The category relationships may include no relationships, parent relationships, and child relationships. A skill may be considered a parent of another skill if the parent skill describes a broader category of skill that includes the child. Other relationships such as grandparent (e.g., a parent's parent), great-grandparent, grandchild, great grandchild and so on may be defined inferentially as well. In some examples, the constructed hierarchy may be organized with broader skills at higher levels and narrower skills at lower levels.

Abstract: A mechanism is provided in a data processing system for partial ingestion of content. The mechanism receives new content to be ingested into a corpus of information. The mechanism applies a plurality of sub-pipelines of annotation engines against the new content in order of effectiveness. The plurality of sub-pipelines include all annotation engines of an ingestion pipeline. Each sub-pipeline within the plurality of sub-pipelines generates one or more intermediate output objects. The mechanism provides access to the one or more intermediate output objects.

Abstract: The present invention relates to a method for comparison of multiple tropical cyclone routes. The method is implementable through a mobile device having an application processor and a touch-sensed display. The method is capable of determining multiple reference cyclone routes relevant to an concerned cyclone route by comparing each of multiple historical cyclone routes, each of which routes consists of multiple cyclone center coordinates, with an estimated cyclone route consisting of multiple estimated cyclone center coordinates for representing the concerned cyclone route.

Abstract: The present description is directed towards systems and methods for directing a user request for content over a network to a given content server on the basis of one or more rules. Methods and systems implemented in accordance with the present description comprise receiving a request for content form a user, the request for content including a profile of the user identifying one or more characteristics associated with the user. One or more rules are retrieved for identifying a content server to which a request for content is to be delivered, the one or more rules including at least one of business rules, network rules, and user profile rules. The one or more retrieved rules are applied to the request for content to identify a content server to which the request for content is to be delivered and the request for content is delivered to the identified content server.

Abstract: A method and a device for rapidly editing sending content based on a receiver. A sending-content-learning module is used to learn sending content to generate a content fragment for a receiver, and perform word frequency statistics on the content fragment; a corresponding list is generated from receiver information, the content fragment, and the word frequency statistics, and is stored; for the receiver and the word frequency statistics, the content fragment is displayed for a sender to edit the sending content, so as to send the sending content to a corresponding receiver. When a user edits a short message or an email, a content fragment commonly used for a corresponding receiver is displayed in a display box, and a sender can rapidly complete editing by selecting a required content fragment, thereby facilitating fast and convenient input and improving the user experience.

Abstract: Systems and methods for aspect-based opinion mining including identifying the polarity (e.g., positive, negative, etc.) of different features of a product or a service as expressed in a text. This general task can be divided into four sub-tasks: identifying the aspect terms, classifying them into one of a set of predefined aspect categories, and identifying the polarity of the aspects terms and the aspect categories. A combination of systems (e.g., rule-based and machine learning based) can be employed to implement aspect category and aspect term polarity mutual annotation for aspect-based sentiment analysis.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for depth concatenation using a matrix computation unit. One of the methods includes: receiving a request to process network inputs to a neural network using an integrated circuit, the neural network comprising a depth concatenation neural network layer; and generating instructions that, when executed by the integrated circuit, cause the integrated circuit to performing operations comprising: for each spatial location in a first input tensor to the depth concatenation layer and a second input tensor to the depth concatenation layer: multiplying, using the matrix computation unit, a second depth vector for the spatial location by a shift weight matrix for the depth concatenation layer to generate a shifted second depth vector; and adding the shifted second depth vector and a first input depth vector for the spatial location to generate a concatenated depth vector.

Abstract: Co-occurrence data representing e.g. preferences and facts observed in a plurality of situations may be stored in a matrix as combinations of high-dimensional sparse vectors. The matrix may be called e.g. as an experience matrix. The data stored in the experience matrix may be subsequently utilized e.g. for predicting a preference of a user in a new situation. A prediction may be determined by a method comprising providing a query comprising one or more query words, accessing the experience matrix containing co-occurrence data stored as vectors of the experience matrix, determining a first auxiliary vector by identifying a vector of the experience matrix associated with a first query word, forming a query vector by using the first auxiliary vector, and determining the prediction by comparing the query vector with the vectors of the experience matrix.

Abstract: Described is a technology by which raw GPS data is processed into segments of a trip, with a predicted mode of transportation (e.g., walking, car, bus, bicycling) determined for each segment. The determined transportation modes may be used to tag the GPS data with transportation mode information, and/or dynamically used. Segments are first characterized as walk segments or non-walk segments based on velocity and/or acceleration. Features corresponding to each of those walk segments or non-walk segments are extracted, and analyzed with an inference model to determine probabilities for the possible modes of transportation for each segment. Post-processing may be used to modify the probabilities based on transitioning considerations with respect to the transportation mode of an adjacent segment. The most probable transportation mode for each segment is selected.

Abstract: Methods, apparatuses and systems directed to pattern identification and pattern recognition. In some particular implementations, the invention provides a flexible pattern recognition platform including pattern recognition engines that can be dynamically adjusted to implement specific pattern recognition configurations for individual pattern recognition applications. In some implementations, the present invention also provides for a partition configuration where knowledge elements can be grouped and pattern recognition operations can be individually configured and arranged to allow for multi-level pattern recognition schemes.

Abstract: Systems, methods, and apparatus for implementing at least one virtual agent. In some embodiments, the at least one virtual agent is programmed to analyze first information regarding a first person; analyze second information regarding a second person different from the first person; and make a joint recommendation for a plurality of persons based at least in part on the first and second information, wherein the plurality of persons comprises the first person and the second person.

Abstract: Methods and systems of localizing a device are presented. In an example method, a communication signal from a device is received by a wireless reference point during a period of time. A sequence of values is generated from the communication signal, as received by the wireless reference point, during the period of time. The sequence of values is supplied to a learning model configured to generate an output based on past values of the sequence of values and at least one predicted future value of the sequence of values. The current location of the device is estimated during the period of time based on the output of the learning model.

Abstract: Provided is a method capable of predicting the quality of cement in a short time period and with high accuracy. The method of predicting the quality or manufacturing conditions of cement through use of a neural network including an input layer and an output layer includes: performing learning of the neural network for a sufficiently large number of times of learning such that ?L<?M is obtained, using learning data and monitor data; then repeating the learning of the neural network until ?L??M is obtained while the number of times of learning is decreased; inputting specific observation data to the input layer of the neural network in which a judgment value for analysis degree obtained from the neural network after the learning is less than a preset value; and outputting an estimated value of specific evaluation data from the output layer of the neural network.

Abstract: A method, devices, computer program products and an internet service is disclosed for adapting a context model. In the method a media clip is received. Also sensor data captured at least partly when the media clip was captured is received. A context is derived using a context model based at least partly on the sensor data and an indication on the context is provided to a user by receiving a search query having at least one context-related search criterion, deriving a similarity between the context and the at least one context-related search criterion and causing at least a part of the media clip to be presented as a response to the search query. Feedback on the relevance of the context is received from the user. Based on the feedback, the context model may then be adapted.

Abstract: A method and system that includes extracting event models from at least one personal planning source of a user, wherein a parameter of an event model includes event location; periodically receiving location information of at least one mobile device of the user; storing the location information in a location log; a pattern worker module maintaining user location patterns through the location log; generating a location prediction from the extracted event models and the user location patterns; a first content worker module checking if the location prediction meets a set of content requirements; if the set of content requirements is satisfied, initiating content retrieval from at least one service; and pushing the content to the mobile device.

Abstract: Briefly, embodiments of methods and/or systems of providing relevant and diverse recommendations are disclosed. For one embodiment, as an example, a system may extract structured and/or semi-structured parameters from web resources obtained from interaction logs comprising records of browsing sessions. Content from extracted parameters may be compared, using an ontology, to find relationships among web resources and query resources.

Abstract: A system for forecast modeling includes at least one processor and at least one database that is operably coupled to the at least one processor. The database includes a time series data module that is configured to store time series data for a domain, an exogenous data module that is configured to store exogenous data associated with multiple exogenous factors and a feedback module that is configured to collect and store feedback data from multiple online users, where the feedback data is related to the exogenous data and the exogenous factors. The system includes a data pre-processor module that is configured to use the at least one processor to identify and select a portion of the exogenous factors using the feedback data collected from the online users for use in a forecast model in combination with the time series data for the domain.

Abstract: A system and method are provided for shared machine learning. The method includes providing a model to a plurality of agents included in a machine learning system. The model specifies attributes and attribute value data types for an event in which the agents act. The method further includes receiving agent-provided inputs during an instance of the event. The agent-provided inputs include estimated attribute values that are consistent with the attribute value data types. The method also includes determining expertise weights for at least some agents in response to at least one ground-truth which is learned from the estimated attribute values. The method additionally includes determining an estimate value for one or more of the attributes using respective adaptive mixtures of the estimated attribute values.

Abstract: Embodiments disclosed herein provide systems, methods, and computer readable media for indicating data object exposure in a cloud computing environment. In a particular embodiment, a method provides receiving information about a data object from the cloud computing environment. The method further provides analyzing the information to determine a plurality of exposure characteristics for the data object. The method further includes determining an indication of exposure of the data object based on the plurality of exposure characteristics.

Abstract: A method and system for matching sequentially relevant content to at least one multimedia content item (MMCI) captured by a mobile device are provided. The method includes extracting at least one MMCI from the mobile device; generating a signature for the extracted at least one MMCI; matching the generated signature to a plurality of signatures of content items; and determining, based on the matching, at least one sequentially relevant content item.