Abstract: A system and method for deploying predictive models is presented which personalizes application program functionality and user interface presentation to the consumer user. That personalization takes place within a decision engine that decouples decision models from application processes so that rapid deployment may take place. When a consumer user specific decision is required, the application program interacts with the decision model to trigger a consumer-user appropriate decision tailored to characteristics of the consumer user. New models are introduced into a decision process without requiring any application program change.

Abstract: Some examples include receiving data from a source external to an adaptive machine learning platform that includes at least one machine learning component. Some implementations may execute a machine learning component to generate a machine learning component output. The machine learning component output may be generated at least in part based on the received data. A command may be generated based at least in part on the machine learning component output, and the command may be communicated to an action plugin. The action plugin manager may be used to configure one or more parameters of an action representing an automated workflow to be executed by the action plugin in response to receiving the command.

Abstract: One or more embodiments of techniques or systems for modeling familiarity for a traveler are provided herein. Familiarity evidence can be received, indicative of how familiar a traveler is with an area or road segment, and based on a number of visits the traveler has made to that area. The familiarity evidence can be used to generate one or more familiarity models indicative of a predicted familiarity of locations around the area. Familiarity models can be based on kernels, graph distances, Markov random fields (MRFs), etc. When route directions are generated from an origin location to a destination location, one or more of the directions can be provided based on one or more of the familiarity models. For example, if a familiarity model indicates that a traveler is familiar with a route, driving directions of the route can be adapted to be more succinct.

Abstract: Methods, systems, and apparatus for generating a model for predicting the characteristics of a user are described. A model template for predicting the one or more characteristics of the selected user is obtained. Training data comprising social relationship information and one or more user characteristics for each of one or more source users is obtained. One or more parameters of the model are determined based on the training data.

Abstract: A model estimation device includes: a data input unit 101; a state number setting unit; an initialization unit; a latent variable variational probability computation unit which computes a variational probability of a latent variable so as to maximize a lower bound of a model posterior probability limited in degree of freedom; a component optimization unit which estimates an optimal type of each component and a parameter thereof so as to maximize the lower bound of the model posterior probability limited in degree of freedom and separated for each component of a latent variable model; a free parameter selection variable computation unit which computes the free parameter selection variable; an optimality determination unit which determines whether or not to continue the maximization of the lower bound of the model posterior probability; and a result output unit.

Abstract: Embodiments relate to providing navigation instructions to a user for reaching a destination via a route planned on a map of a navigation system, said map comprising a plurality of road segments, at least a subset of said road segments comprising an assigned familiarity score. An aspect includes tracking a route taken by a user. Another aspect includes calculating an updated familiarity score of the one or more road segments of the tracked route using at least one of a user-defined updating variable and an updating variable based on at least one dynamically obtained actual travelling condition. Yet another aspect includes storing the updated familiarity score in a data storage of the navigation system.

Abstract: In an exemplary embodiment, a computer-implemented method includes receiving an instruction to select an output set from an input set, where the output set is a top subset or a bottom subset of the input set, where the input set comprises a plurality of members, and where each member of the input set includes a plurality of bits. A first subset of the plurality of bits is selected. A histogram is generated, by a computer processor, based on the values in the first subset of the plurality of bits. A threshold value of the input set is determined, where the threshold value separates the values of the output set from the values of the remainder of the input set, and where the threshold value is based at least in part on the histogram. The output set is then extracted from the input set based on the threshold value.

Abstract: An autonomous vehicle configured to avoid pedestrians using hierarchical cylindrical features. An example method involves: (a) receiving, at a computing device, range data corresponding to an environment of a vehicle, and the range data comprises a plurality of data points; (b) detecting, by the computing device, one or more subsets of data points from the plurality of data points that are indicative of an upper-body region of a pedestrian, and the upper-body region may comprise parameters corresponding to one or more of a head and a chest of the pedestrian; and (c) in response to detecting the one or more subsets of data points, determining a position of the pedestrian relative to the vehicle.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training machine learning models. The models can include models for predicting a next transaction price or a next transaction price direction for one or more financial products, for classifying particular debit or credit card transactions as likely being anomalous or fraudulent or not, or for classifying particular financial claims processing transactions, e.g., insurance, health care, or employee expense claims transactions, as likely being anomalous or fraudulent or not.

Abstract: A method of providing a semantic data architecture includes providing a data model layer. The data model layer is formed by a processor, a storage device, a memory, and a communication interface in combination with a data model application program stored in the storage device. The processor, running the data model application program, is configured to use the storage device, memory, and communication interface to selectively receive source data from a source device, process the corresponding source data based on pre-defined data types and filtering terms to form semantic data arranged in a binary tree structure, and store the semantic data in the storage device. The method also includes providing a data filtering layer. The method may also include providing memory model, general purpose parser, backward inference, primitive functions, rewriting engine, and reasoning engine layers in various combinations. An apparatus for providing a semantic data architecture is also provided.

Abstract: A system and an article of manufacture for semantic and contextual searching over a knowledge repository including creating a search query for each concept related to the target concept to form a search context, wherein the search query for each related concept comprises at least one word derived from a record of that concept previously authored in the project, running the search query on a search index of a knowledge repository to identify a record of the related concept for which the search query is created, and fetching the record of the target concept from the repository as a search result such that the fetched record of the target concept is linked in the knowledge repository to a record of the related concept returned as a result of running the search query on at least one record of the at least one related concept.

Abstract: A method for semantic and contextual searching over a knowledge repository. The method includes creating a search query for each concept related to the target concept to form a search context, wherein the search query for each related concept comprises at least one word derived from a record of that concept previously authored in the project, running the search query on a search index of a knowledge repository to identify a record of the related concept for which the search query is created, and fetching the record of the target concept from the repository as a search result such that the fetched record of the target concept is linked in the knowledge repository to a record of the related concept returned as a result of running the search query on at least one record of the at least one related concept.

Abstract: Apparatus and methods for arbitration of control signals for robotic devices. A robotic device may comprise an adaptive controller comprising a plurality of predictors configured to provide multiple predicted control signals based on one or more of the teaching input, sensory input, and/or performance. The predicted control signals may be configured to cause two or more actions that may be in conflict with one another and/or utilize a shared resource. An arbitrator may be employed to select one of the actions. The selection process may utilize a WTA, reinforcement, and/or supervisory mechanisms in order to inhibit one or more predicted signals. The arbitrator output may comprise target state information that may be provided to the predictor block. Prior to arbitration, the predicted control signals may be combined with inputs provided by an external control entity in order to reduce learning time.

Abstract: A device executes a virtual interview application, and receives first user input, via the virtual interview application, to create multiple first media clips for a first virtual interview, with each of the first media clips including a different interview question. The device submits, from the device to a remote network device, the multiple first media clips for conducting the first virtual interview with a plurality of first interviewees.

Abstract: Machines, systems and methods for classifying documents, the method comprising: classifying a document from among a plurality of documents in a first class, in response to applying statistical analysis to data associated with the document; classifying the document in a second class, in response to determining that a rule from among a plurality of rules applies to the document, wherein a proposed rule is added to the plurality of rules, in response to determining that application of the proposed rule to one or more of the plurality of documents to which the rule is applicable does not diminish accuracy of overall classification for the plurality of documents.

Abstract: Technology for policies with reduced associated costs is disclosed. A policy may include an ordered rule set. When evaluated, the highest priority rule in the order that does not skip may control the policy outcome. Rules within a policy may have associated costs, such as data fetch and evaluation costs. In some contexts, it may be less important to evaluate every rule than to evaluate the policy quickly. Reduced policies that have one or more rules removed or that skip evaluation of some rules may be created for these contexts. When a rule of a policy is skipped, it may result in a possibility of a false allow or false deny. In some cases, rules may be duplicative. Removal or skipping of duplicative rules does not increase the possibility of a false allow or false deny. By using reduced policies in identified contexts, policy evaluation costs may be reduced.

Abstract: A computer-implemented method for object recognition using a recursive cortical network comprising receiving an input image at an input module, applying a trained recursive cortical network (RCN) to the image using an inference module to activate child features of the RCN, selecting pools of the RCN containing the activated child features, propagating the selection of the pools to identify probabilities of one or more high-level features matching one or more objects in the input image.

Abstract: A low-density parity check min-sum decoder including a variable node processing unit having N+1 inputs. A first bank of N+1 two-input adders each have an associated output, and at least one of the N+1 inputs go to more than two of the adders of the first bank. A second bank of N two-input adders has no adders in common with the first bank. At least one of the adders of the first bank provides its associated output to more than one adder of the second bank. The banks of adders are disposed in series. A sign module outputs a sign value produced from one of the inputs and an output from one of the adders of the second bank. N+1 outputs are provided, where one of the outputs is the sign value.

Abstract: A method for estimating artist ambiguity in a dataset is performed at a device with a processor and memory storing instructions for execution by the processor. The method includes applying a statistical classifier to a first dataset including a plurality of media items, wherein each media item is associated with one of a plurality of artist identifiers, each artist identifier identifies a real world artist, and the statistical classifier calculates a respective probability that each respective artist identifier is associated with media items from two or more different real world artists based on a respective feature vector corresponding to the respective artist identifier. The method further includes providing a report of the first dataset, including the calculated probabilities, to a user of the electronic device. Each respective feature vector includes a plurality of features that indicate likelihood of artist ambiguity.

Abstract: An influence metric describing the influence of a social networking system object on social networking system users is determined based on affinities between the users and the object. For example, affinities between the associated users and the object are combined to determine the influence metric. Content may be selected for presentation to users based in part on influence metrics of the content. Additionally, influence metrics of objects associated with a user may be combined to determine the relevance of objects associated with the user, which may also be used to select content for presentation to the user.

Abstract: A method of evaluating a value of reference information and an apparatus for evaluating a value of reference information are disclosed. According to an embodiment of the present invention, the method includes calculating similarity values of reference relationships between the subject document and each of neighbor documents, respectively, and summing the calculated similarity values to calculate a first sum, multiplying each of the similarity values of reference relationships between the subject document and each of the neighbor documents by a corresponding value of reference information of the first reference document for each of the neighbor documents, and summing the multiplied values to calculate a second sum, and evaluating a value of reference information using a ratio of the first and second sums.

Abstract: Apparatuses, systems, methods, and computer program products are disclosed for machine learning results. An input module may receive user input identifying a value for a machine learning parameter. A display module may display one or more machine learning results for the identified machine learning parameter in response to the input module receiving the user input. An update module may dynamically update the displayed one or more machine learning results in response to the input module receiving additional user input identifying an additional value for the machine learning parameter. A pre-compute module may predetermine permutations of the machine learning results prior to the input module receiving the user input.

Abstract: Certain aspects of the present disclosure support a technique for strategic synaptic failure and learning in spiking neural networks. A synaptic weight for a synaptic connection between a pre-synaptic neuron and a post-synaptic neuron can be first determined (e.g., according to a learning rule). Then, one or more failures of the synaptic connection can be determined based on a set of characteristics of the synaptic connection. The one or more failures can be omitted from computation of a neuronal behavior of the post-synaptic neuron.

Abstract: A resource allocation device for cloud computing is disclosed. The resource allocation device may include a resource information collecting unit to collect information on tasks to be processed and resource used for the tasks; and a resource allocation unit to allocate the tasks to the resources using initial value setting and cycle deletion (IVSCD) according to the information so that utility is increased.

Abstract: Image analysis includes: determining, using one or more processors, an image quality score associated with an image, including: determining a foreground and a background in the image; calculating a set of one or more characteristic parameters of the image based on the determined foreground and background; calculating the image quality score based at least in part on the set of characteristic parameters, wherein calculating the image quality score comprises using an image quality computation model that has been pre-trained; and in response to a search query, generating a set of search results that includes a set the images, wherein inclusion of the images or ranking of the search results is based at least in part on image quality scores associated with the set of images.

Abstract: Disclosed are methods, systems, and articles of manufactures for providing evolving information in generating a physical design with custom conductivity using force models and design space decomposition by first presenting a layout area in an interface. The interface then displays the evolution of the physical design in the interface to reflect temporal states of the physical design during generation of the physical design after the system receives an input for the physical design and a request for creating the physical design.

Abstract: Disclosed are methods and systems of creating, evaluating, and tuning a predictive engine for machine learning, including steps to deploy the predictive engine with an initial parameter set; receive queries to the deployed engine variant and in response, generate predicted results; receive corresponding actual results; associate the queries, the predicted results, and the actual results with a replay tag; evaluate the performance of the deployed engine variant; generate a new engine parameter set based on tuning of one or more parameters of the initial engine parameter set, according to the evaluation results; deploy the new engine variant to replace the initial engine variant; receive a replay request from an operator specifying the currently or a previously deployed engine variant; and in response to the replay request, replay at least one of the queries, the corresponding predicted results, the actual results, and the evaluation results.

Abstract: A system and method for fusing the outputs from multiple LiDAR sensors on a vehicle that includes cueing the fusion process in response to an object being detected by a radar sensor and/or a vision system. The method includes providing object files for objects detected by the LiDAR sensors at a previous sample time, where the object files identify the position, orientation and velocity of the detected objects. The method projects object models in the object files from the previous sample time to provide predicted object models. The method also includes receiving a plurality of scan returns from objects detected in the field-of-view of the sensors at a current sample time and constructing a point cloud from the scan returns. The method then segments the scan points in the point cloud into predicted scan clusters, where each cluster identifies an object detected by the sensors.

Abstract: A highly reliable management apparatus and management method capable of allocating a resource that satisfies a user's request is suggested.

Abstract: Various embodiments of the invention provide methods, systems, and computer program products for determining a pattern for communications conducted between agents at a contact center and contact parties. In particular embodiments, the pattern is based on a utilization of one or more information resources by the agents during the communications. Depending on the embodiment, the communications may comprise one or more of telephone calls, text messages, emails, and Web chats. In addition, in various embodiments, a particular information resource from among the one or more information resources is identified based on the pattern and is associated with a keyphrase found in at least two of the communications. Accordingly, as a result of the association, the particular information resource is made available to at least one agent at the contact center during a time the agent is fielding a subsequent communication in which the keyphrase is detected.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for obtaining a plurality of model representations of predictive models, each model representation associated with a respective user and expresses a respective predictive model, and selecting a model implementation for each of the model representations based on one or more system usage properties associated with the user associated with the corresponding model representation.

Abstract: The present invention relates generally to a system and method for the reordering of text predictions. More particularly, the system and method reorders the text predictions based on modified probability values, wherein the probability values are modified according to the likelihood that a given text prediction will occur in the text inputted by a user.

Abstract: An apparatus includes a memory and a processor. The memory may be configured to store at least a portion of a multi-level tree representation of an ordered multi-field rule-based classification list. The tree representation includes at least one non-leaf level and one or more leaf levels. Each entry in the at least one non-leaf level contains a count value indicating a number of rules having a matching field. Entries in at least one of the one or more leaf levels include rule pointers arranged in priority order. The processor may be configured to incrementally insert or delete rules, while preserving ordering semantics of the tree representation.

Abstract: Methods, systems, and computer-readable media for providing topical search suggestions are provided. Topical search suggestions allow a user to receive search results related to the designated topic or subject matter. The present invention may generate multiple topics based on search input provided by a user. The search input may be a search prefix that includes one or more words entered into the search query box before the completed search query is submitted to the search engine. A search interface then presents the topics derived from the search prefix to a user before the user submits the query. In another embodiment, the user designates multiple search inputs. The present invention generates search results based on the search inputs and then presents topics extracted from the search results. In one embodiment, the topics are extracted by performing a natural language analysis of search result metadata.

Abstract: In one embodiment, a plurality of objects associated with a dataset and a specified number of nearest objects to be identified are received. The received objects are sorted in a structured format. Further, a key object and a number of adjacent objects corresponding to the key object are selected from the sorted plurality of objects, wherein the number of adjacent objects is selected based on the specified number of nearest objects to be identified. Furthermore, distances between the key object and the number of adjacent objects are determined to identify the specified number of nearest objects, wherein the distances are determined until the specified number of nearest objects is identified. Based on the determined distances, the specified number of nearest objects in the dataset is identified for data analysis.

Abstract: An apparatus, system and method for automatically evaluating a transaction request are provided. An adaptive modeling platform builds models and deploys them in a very systematic manner, without requiring much or any human intervention. This system ensures an end to end data management; encompassing variable generation, model building and evaluation of the built models, decision logic and strategy design. It guarantees deployment of these predictive models in real time, monitors performance of the portfolio and generates reports & alerts for the same. The system periodically examines the models in production and rebuilds them when their performance falls below a predefined threshold. When the human discretion so permits, the system can be interrupted at any point in time and changes can be made wherever desired in the process.

Abstract: The disclosure relates to a communication client application for running on a user terminal to conduct calls over a network. The client is configured to access a model which models quality of user experience for calls based on a set of technical parameters of each call. The model itself is based on user feedback indicating subjective quality of multiple past calls as experienced by multiple users, modeled with respect to the technical parameters collected from each of the past calls. The model generates a predicted call quality score predicting the quality of user experience for the call given its technical parameters. Based on this process, one or more of the technical parameters of the call can be adapted to try to increase the quality experienced by the user.

Abstract: A system includes a free itemset generation unit to generate a set of free itemsets, each having a frequency in the database greater than or equal to a predetermined threshold value set in advance, a valid rule candidate generation unit to generate rule candidates and store the generated rule candidates, and a rule minimality decision unit to check minimality of each of the generated rule candidates and output the generated rule candidate to an output apparatus when the generated rule candidate is determined to be minimal.

Abstract: The current application is directed to monitoring subsystems, and monitoring methods incorporated within the monitoring subsystems, that monitor operation of devices and systems in order to identify normal states and to quickly determine when a device or system transitions from a normal state to an abnormal state. The methods and monitoring components to which the current application is directed employ self-organizing maps and moving-average self-organizing maps to both characterize normal system behavior and to identify transitions to abnormal system behaviors.

Abstract: A method of identifying one or more particular images from an image collection, includes indexing the image collection to provide image descriptors for each image in the image collection such that each image is described by one or more of the image descriptors; receiving a query from a user specifying at least one keyword for an image search; and using the keyword(s) to search a second collection of tagged images to identify co-occurrence keywords. The method further includes using the identified co-occurrence keywords to provide an expanded list of keywords; using the expanded list of keywords to search the image descriptors to identify a set of candidate images satisfying the keywords; grouping the set of candidate images according to at least one of the image descriptors, and selecting one or more representative images from each grouping; and displaying the representative images to the user.

Abstract: A method, system and computer program product for generating answers to questions. In one embodiment, the method comprises receiving an input query, decomposing the input query into a plurality of different subqueries, and conducting a search in one or more data sources to identify at least one candidate answer to each of the subqueries. A ranking function is applied to each of the candidate answers to determine a ranking for each of these candidate answers; and for each of the subqueries, one of the candidate answers to the subquery is selected based on this ranking. A logical synthesis component is applied to synthesize a candidate answer for the input query from the selected the candidate answers to the subqueries. In one embodiment, the procedure applied by the logical synthesis component to synthesize the candidate answer for the input query is determined from the input query.

Abstract: Systems and methods for searching domain names and owner information associated with the domain names in a web-based Whols search tool. Systems including one or more databases and one or more text indexes may store data including the domain names and owner information, and index text-searchable portions of the data. A search request may be received from a user with user preferences such as a selection of multiple or all top level domains. Search results may be ranked by relevancy, and may include the domain name with associated owner information.

Abstract: A particular smart hazard detector may itself function as a guide during a process of installation of the same at an installation location. Additionally, the installation location of the particular smart hazard detector may play a central role in how various settings of the smart hazard detector are defined and adjusted over time.

Abstract: Systems and methods to generate modified/refined user inputs based on the original user input, such as a search query, are disclosed. The method may be implemented for Roman-based and/or non-Roman based language such as Chinese. The method may generally include receiving an original user input and identifying core terms therein, determining potential alternative inputs by replacing core term(s) in the original input with another term according to a similarity matrix and/or substituting a word sequence in the original input with another word sequence according to an expansion/contraction table where one word sequence is a substring of the other, computing likelihood of each potential alternative input, and selecting most likely alternative inputs according to a predetermined criteria, e.g., likelihood of the alternative input being at least that of the original input. A cache containing pre-computed original user inputs and corresponding alternative inputs may be provided.

Abstract: Methods, apparatus, and articles of manufacture to measure search results are disclosed. A disclosed example method to measure search results includes identifying a preview event for a search result associated with a search query, and storing the preview event in association with a search engine identifier and a web page identifier.

Abstract: A system and method for filtering an already obtained information resource (“document”) for display is described, including identifying a target portion of the information resource; dividing the data of the target portion into elements of a set; receiving one or more search requests as a filter specification; interpreting the filter specification as a Boolean logic expression; evaluating the elements of the set based on the Boolean logic expression, in an ongoing manner, each time the filter specification is entered or modified; and dynamically re-rendering the display, upon each such entry or modification of the filter specification, with all elements not matching the logic expression removed from the display of the target portion, providing the user with a “shrinking document” more likely to contain just the information being looked for.

Abstract: A method and associated system for automatically deriving context of a business rule that is being inferred from an information source, such as legacy computer-program source code, a system specification, or a description of a business function or operation. A first generation of context-dependent conditions upon the business rule is identified by analyzing the information source and other, related, sources, and the business rule is modified to account for this first generation of context-dependent conditions. Context-dependent conditions that place constraints upon the previously identified first-generation conditions are then identified and the business rule is further modified to account for this next generation of context-dependent conditions. This process repeats iteratively through multiple generations of identifications, until the business rule has been modified to accommodate all contextual information that may be gleaned from all analyzed information sources.

Abstract: Techniques are presented for providing a software fitting assessment. The techniques may be performed by methods, apparatus, and/or computer program products. The techniques include automatically matching on a computer system one or more specified requirements for a project with one or more software functions stored in a repository. The automatically matching includes mining the repository in order to match requirements. The repository includes software functions, requirements accumulated from previous projects, and results of stored matches between the software functions and the requirements accumulated from previous projects. The techniques include outputting by the computer system one or more results of the matching.

Abstract: Systems and methods are disclosed for active column filtering. In accordance with one implementation, a method is provided for active column filtering. The method includes providing a table having data values arranged in rows and columns, providing a first filter location indicator whose location is visually associated with a first column, and providing a first interface based on a selection of the first filter location indicator, wherein the first interface's location is visually associated with the first column. The method also includes acquiring a first filter input entered into the first interface, filtering the table based on the acquired first filter input, providing the filtered table for displaying, and providing an applied filter indicator, whose location is visually associated with the first column, the applied filter indicator including at least the first filter input.

Abstract: A system for executing services in a distributed way, the execution of each service involving the processing of a plurality of tasks of predetermined types, each task type to be processed being represented by a corresponding type of tuples of data relating to the task type, wherein: the system includes a plurality of peer entities; each peer entity being adapted to store and execute a plurality of worker modules, each worker module being adapted to process at least one task type; each peer entity including tuple handling modules that, for a tuple of a type T, are adapted: a) to check if at least one of the plurality of the worker modules in the peer entity is adapted to process the task represented by the tuple, b) in the positive case of a), to take charge of the processing of the task represented by the tuple, and c) in the negative case of a), to execute a forwarding procedure for forwarding the tuple to another peer entity of the plurality of peer entities of the system.