Abstract: A first user and a second user may exchange messages during a session and the exchanged messages may be processed to suggest resources to the first user. For example, the first user may be a customer service representative assisting a customer, and the suggested resources may include text of a message to send to the customer. The resources may be suggested by computing a semantic representation of the messages in the session, computing a context vector that describes a context of the session, computing a context hash vector from the context vector, and retrieving one or more resources from a data store of resources using the context hash vector. Information about a resource may be presented to the first user to assist the first user in responding to the second user.

Abstract: Aspects of the disclosure generally relate to computing devices and may be generally directed to devices, systems, methods, and/or applications for learning conversations among two or more conversation participants, storing this knowledge in a knowledgebase (i.e. neural network, graph, sequences, etc.), and enabling a user to simulate a conversation with an artificially intelligent conversation participant.

Abstract: A system, method, and computer program product for automatically selecting from a plurality of analytic algorithms a best performing analytic algorithm to apply to a dataset is provided. The automatically selecting from the plurality of analytic algorithms the best performing analytic algorithm to apply to the dataset enables a training a plurality of analytic algorithms on a plurality of subsets of the dataset. Then, a corresponding prediction accuracy trend is estimated across the subsets for each of the plurality of analytic algorithms to produce a plurality of accuracy trends. Next, the best performing analytic algorithm is selected and outputted from the plurality of analytic algorithms based on the corresponding prediction accuracy trend with a highest value from the plurality of accuracy trends.

Abstract: A mechanism is provided in a data processing system for question answering using context features in closure form. The mechanism receives a function call comprising an input question and a set of non-local context evidence in closure form. The mechanism decomposes the input question into one or more queries and applies the one or more queries to a corpus of information to obtain a set of local hypothesis evidence. The mechanism generates hypotheses for answering the input question based on the local hypothesis evidence and the set of non-local context evidence.

Abstract: The technology disclosed provides a so-called “pointer sentinel mixture architecture” for neural network sequence models that has the ability to either reproduce a token from a recent context or produce a token from a predefined vocabulary. In one implementation, a pointer sentinel-LSTM architecture achieves state of the art language modeling performance of 70.9 perplexity on the Penn Treebank dataset, while using far fewer parameters than a standard softmax LSTM.

Abstract: A system and method is described herein for performing at least one industrial process at each facility of a plurality of facilities based on an industrial process standard generated by reducing functional, and trend line outlier bias in data of one or more process parameters as measured by one or more sensors. Outliers are removed from the data set through an objective method. Bias is determined based on absolute, relative error, or both. Error values are computed from the data, model coefficients, or trend line estimates. Outlier data records are removed when the error values are greater than or equal to one or more criteria.

Abstract: In one embodiment, a method includes accessing a first set of objects associated with an online social network, each object being associated with one or more comments. The method also includes generating a second set of objects from the first set of objects by applying a first filtering criteria to the first set of objects and scoring each object in the second set of objects based on the comments associated with each object. The method further includes generating a training set of objects from the second set of objects by selecting each object from the second set of objects having a score greater than a first threshold score, each object in the training set being associated with a first object-classification. The method further includes determining an object-classifier algorithm for the first object-classification, the object-classifier algorithm being determined through an iterative training process performed one or more times.

Abstract: A technology is described for determining a launch workflow for launching multiple computing instances on a physical host in a computing service environment using a predicted launch time. An example method may include receiving a launch plan to launch a plurality of computing instances on a physical host within a computing service environment. A first launch workflow and a second launch workflow may then be determined for the launch plan where the first launch workflow and the second launch workflow may specify different sequences of operations performed to launch the computing instance on the physical host. Predicted launch times for the first launch workflow and the second launch workflow may be obtained and the predicted launch times may be compared to determine a launch workflow used in executing the launch plan.

Abstract: Systems and methods for preventing machine learning models from negatively affecting mobile devices are provided. For example, a mobile device including a camera, memory devices, and one or more processors are provided. In some embodiments, the processors may be configured to provide images captured by the camera to a machine learning model at a first rate. The processors may also be configured to determine whether one or more of the images includes an object. If one or more of the images includes the object, the processors may be further configured to adjust the first rate of providing the images to the machine learning model to a second rate, and in some embodiments, determine whether to adjust the second rate of providing the images to the machine learning model to a third rate based on output received from the machine learning model.

Abstract: Systems and methods (e.g., utilities) for use in providing automated data quality detection that can be used multiple times across various domains and across multiple data quality spheres. A data structure or schema of an incoming data set is initially mapped to a desired data or knowledge state in a domain ontology made up of a number of TBox statements. Data quality issues in the incoming data set can then be detected as instances of the incoming data set are attempted to be inferenced against or otherwise matched to anticipated TBox statements of the domain ontology.

Abstract: Techniques and systems are provided for predictive modeling based on interactions with a network device. For example, a method may include generating a prediction including a correlation between an interaction with a network device and a context, wherein the interaction is associated with a function performed by the network device. Confidence parameters associated with the prediction can be determined. The prediction can be tested by analyzing received interaction data and contextual data, and the analysis can include determining whether the interaction with the network device occurred in the correlated context. A confidence value can be calculated based on the testing outcome, and can be compared to the confidence parameters. A message relating to modification of the confidence parameters can be transmitted.

Abstract: A system includes a centralized repository for tracking rule content and managing subscriptions to rule content by organizations and providers utilizing the system; a rule-evaluation server for receiving requests for rule-evaluations for specific patients, wherein the server determines content needing to be evaluated and retrieves the content to be used; a rule engine for performing the evaluations, wherein content, patient data, and rule evaluation parameters are provided to the engine, and the engine returns recommendations triggered by the evaluation, if any; an aggregator for aggregating recommendations from multiple sources, detecting and coordinating related recommendations, and applying configuration settings based on the patient and/or provider in context; and a client component for coordinating communication between an electronic health records system, the server, and the aggregator.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for evaluating reinforcement learning policies. One of the methods includes receiving a plurality of training histories for a reinforcement learning agent; determining a total reward for each training observation in the training histories; partitioning the training observations into a plurality of partitions; determining, for each partition and from the partitioned training observations, a probability that the reinforcement learning agent will receive the total reward for the partition if the reinforcement learning agent performs the action for the partition in response to receiving the current observation; determining, from the probabilities and for each total reward, a respective estimated value of performing each action in response to receiving the current observation; and selecting an action from the pre-determined set of actions from the estimated values in accordance with an action selection policy.

Abstract: Disclosed herein is a computer implemented method and system for analyzing load responsive behavior of infrastructure components in an electronic environment for proactive management of the infrastructure components. Transaction data on multiple application transactions is collected. Load patterns are identified from the collected transaction data for generating load profiles. Data on infrastructure behavior in response to the application transactions is collected. Infrastructure behavior patterns are identified from the infrastructure behavior data for generating behavior profiles. The generated load profiles and the generated behavior profiles are correlated to create a load responsive behavior model. The created load responsive behavior model predicts behavior of the infrastructure components for different load patterns. A live data stream from current application transactions is analyzed using the load responsive behavior model to determine current load responsive behavior.

Abstract: A method comprising using at least one hardware processor for receiving a topic under consideration (TUC); providing the TUC as input to a claim function, wherein the claim function is configured to mine at least one content resource, and applying the claim function to the at least one content resource, to extract claims with respect to the TUC; and providing the TUC as input to a classification function, and applying the classification function to one or more claims of the extracted claims, to output corresponding one or more classification tags, wherein each classification tag is associated with its corresponding claim.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for content recommendation using neural networks. One of the methods includes receiving context information for an action recommendation; processing the context information using a neural network that comprises one or more Bayesian neural network layers to generate, for each of the actions, one or more parameters of a distribution over possible action scores for the action and selecting an action from plurality of possible actions using the parameters of the distributions over the possible action scores for the action.

Abstract: A neural network includes inputs for receiving input signals, and synapses connected to the inputs and having corrective weights organized in an array. Training images are either received by the inputs as an array or codified as such during training of the network. The network also includes neurons, each having an output connected with at least one input via one synapse and generating a neuron sum array by summing corrective weights selected from each synapse connected to the respective neuron. Furthermore, the network includes a controller that receives desired images in an array, determines a deviation of the neuron sum array from the desired output value array, and generates a deviation array. The controller modifies the corrective weight array using the deviation array. Adding up the modified corrective weights to determine the neuron sum array reduces the subject deviation and generates a trained corrective weight array for concurrent network training.

Abstract: The invention proposes that a software agent for assisting a user in a virtual environment defined by a first concept map, maintains a record of its experiences in the form of a second concept map. The concept maps are compared to obtain a measure of stimulation. The measure of stimulation is used to derive a comparison value indicative an emotional state of the user, and the comparison value is used in a reward function to guide the behavior of the software agent.

Abstract: A computer-implemented content suggestion engine provides content suggestions to a requesting user based on information about content items that other users may have independently categorized or organized into folders within a content repository. Embodiments of the method comprise a content repository having a plurality of content items, where each content item is associated with one or more user-created folders. Embodiments further comprise receiving, via a network, a suggestion request for suggested content, where the suggestion request identifies a first content item for which suggestions are sought. Other content items in the content repository are then identified as potential suggestions based on the application of a formal relationship between the first content item and the potential suggested content items. One or more of the potential suggested content items may then be provided in response to the suggestion request via the network.

Abstract: A user management method according to the present disclosure includes: storing appliance use information including: appliance identification information for identifying an appliance; user information for identifying a user of the appliance; and an operating state of the appliance when the appliance was used; analyzing the appliance use information stored in the storing to identify, from among a plurality of functions of the appliance, one or more first functions each having a use frequency less than or equal to a threshold value; and providing the user with a notice which prompts use of the one or more first functions identified in the analyzing.