Abstract: In some examples, a computing device refines feature information of query text. The device repeatedly determines attention information based at least in part on feature information of the image and the feature information of the query text, and modifies the feature information of the query text based at least in part on the attention information. The device selects at least one of a predetermined plurality of outputs based at least in part on the refined feature information of the query text. In some examples, the device operates a convolutional computational model to determine feature information of the image. The device network computational models (NCMs) to determine feature information of the query and to determine attention information based at least in part on the feature information of the image and the feature information of the query. Examples include a microphone to detect audio corresponding to the query text.

Abstract: A processing unit can determine a first feature value corresponding to a session by operating a first network computational model (NCM) based part on information of the session. The processing unit can determine respective second feature values corresponding to individual actions of a plurality of actions by operating a second NCM. The second NCM can use a common set of parameters in determining the second feature values. The processing unit can determine respective expectation values of some of the actions of the plurality of actions based on the first feature value and the respective second feature values. The processing unit can select a first action of the plurality of actions based on at least one of the expectation values. In some examples, the processing unit can operate an NCM to determine expectation values based on information of a session and information of respective actions.

Abstract: Systems and methods for determining knowledge-guided information for a recurrent neural networks (RNN) to guide the RNN in semantic tagging of an input phrase are presented. A knowledge encoding module of a Knowledge-Guided Structural Attention Process (K-SAP) receives an input phrase and, in conjunction with additional sub-components or cooperative components generates a knowledge-guided vector that is provided with the input phrase to the RNN for linguistic semantic tagging. Generating the knowledge-guided vector comprises at least parsing the input phrase and generating a corresponding hierarchical linguistic structure comprising one or more discrete sub-structures. The sub-structures may be encoded into vectors along with attention weighting identifying those sub-structures that have greater importance in determining the semantic meaning of the input phrase.

Abstract: This document relates to architectures and training procedures for multi-task machine learning models, such as neural networks. One example method involves providing a multi-task machine learning model having one or more shared layers and two or more task-specific layers. The method can also involve performing a pretraining stage on the one or more shared layers using one or more unsupervised prediction tasks.

Abstract: A workflow engine tool is disclosed that enables scientists and engineers to programmatically author workflows (e.g., a directed acyclic graph, “DAG”) with nearly no overhead, using a simpler script that needs almost no modifications for portability among multiple different workflow engines. This permits users to focus on the business logic of the project, avoiding the distracting tedious overhead related to workflow management (such as uploading modules, drawing edges, setting parameters, and other tasks). The workflow engine tool provides an abstraction layer on top of workflow engines, introducing a binding function that converts a programming language function (e.g., a normal python function) into a workflow module definition. The workflow engine tool infers module instances and induces edge dependencies automatically by inferring from a programming language script to build a DAG.

Abstract: Systems and methods are disclosed for intelligent scheduling of calls to sales leads, leveraging machine learning (ML) to optimize expected results. One exemplary method includes determining, using a connectivity prediction model, call connectivity rate predictions; determining timeslot resources; allocating, based at least on the call connectivity rate predictions and timeslot resources, leads to timeslots in a first time period; determining, within a timeslot and using a lead scoring model, lead prioritization among leads within the timeslot; configuring, based at least on the lead prioritization, the telephone unit with lead information for placing a phone call; and applying a contextual bandit (ML) process to update the connectivity prediction model, the lead scoring model, or both. During subsequent time periods, the updated connectivity prediction and lead scoring models are used, thereby improving expected results over time.

Abstract: A workflow engine tool is disclosed that enables scientists and engineers to programmatically author workflows (e.g., a directed acyclic graph, “DAG”) with nearly no overhead, using a simpler script that needs almost no modifications for portability among multiple different workflow engines. This permits users to focus on the business logic of the project, avoiding the distracting tedious overhead related to workflow management (such as uploading modules, drawing edges, setting parameters, and other tasks). The workflow engine tool provides an abstraction layer on top of workflow engines, introducing a binding function that converts a programming language function (e.g., a normal python function) into a workflow module definition. The workflow engine tool infers module instances and induces edge dependencies automatically by inferring from a programming language script to build a DAG.

Abstract: Techniques of selective surfacing email messages in computing systems are disclosed herein. In one embodiment, a method includes calculating an importance value corresponding to an incoming email based on one or more values of attributes of the incoming email and corresponding contribution values of the features toward the importance value. The method can then include performing a comparison between the calculated importance value of the incoming email and a preset importance threshold and, based on the performed comparison, selectively surfacing the incoming email to the user in a first section of an email inbox irrespective of a date/time of reception of the incoming email relative to other emails in the inbox.

Abstract: A processing unit can determine multiple representations associated with a statement, e.g., subject or predicate representations. In some examples, the representations can lack representation of semantics of the statement. The computing device can determine a computational model of the statement based at least in part on the representations. The computing device can receive a query, e.g., via a communications interface. The computing device can determine at least one query representation, e.g., a subject, predicate, or entity representation. The computing device can then operate the model using the query representation to provide a model output. The model output can represent a relationship between the query representations and information in the model. The computing device can, e.g., transmit an indication of the model output via the communications interface.

Abstract: Systems, methods, and computer-executable instructions for approximating a softmax layer are disclosed. A small world graph that includes a plurality of nodes is constructed for a vocabulary of a natural language model. A context vector is transformed. The small world graph is searched using the transformed context vector to identify a top-K hypothesis. A distance from the context vector for each of the top-K hypothesis is determined. The distance is transformed to an original inner product space. A softmax distribution is computed for the softmax layer over the inner product space of the top-K hypothesis. The softmax layer is useful for determining a next word in a speech recognition or machine translation.

Abstract: Generally discussed herein are devices, systems, and methods for scheduling tasks to be completed by resources. A method can include identifying features of the task, the features including a time-dependent feature and a time-independent feature, the time-dependent feature indicating a time the task is more likely to be successfully completed by the resource, converting the features to feature values based on a predefined mapping of features to feature values in a first memory device, determining, by a gradient boost tree model and based on a first current time and the feature values, a likelihood the resource will successfully complete the task, and scheduling the task to be performed by the resource based on the determined likelihood.

Abstract: Generally discussed herein are devices, systems, and methods for task routing. A method can include receiving, from a resource, a request for a task, in response to receiving the request, determining whether to retrieve a new task of new tasks stored in a first queue or a backlog task of backlog tasks stored in a second queue based on a combined amount of backlog tasks and new tasks relative to a capacity of the resource or the resources, retrieving the new task or the backlog task from the determined first queue or second queue, respectively, based on the determination, and providing the retrieved task to the resource.

Abstract: Described herein are systems, methods, and techniques by which a processing unit can build an end-to-end dialogue agent model for end-to-end learning of dialogue agents for information access and apply the end-to-end dialogue agent model with soft attention over knowledge base entries to make the dialogue system differentiable. In various examples the processing unit can apply the end-to-end dialogue agent model to a source of input, fill slots for output from the knowledge base entries, induce a posterior distribution over the entities in a knowledge base or induce a posterior distribution of a target of the requesting user over entities from a knowledge base, develop an end-to-end differentiable model of a dialogue agent, use supervised and/or imitation learning to initialize network parameters, calculate a modified version of an episodic algorithm. e.g., the REINFORCE algorithm, for training an end-to-end differentiable model based on user feedback.

Abstract: Examples are generally directed towards context-sensitive generation of conversational responses. Context-message-response n-tuples are extracted from at least one source of conversational data to generate a set of training context-message-response n-tuples. A response generation engine is trained on the set of training context-message-response n-tuples. The trained response generation engine automatically generates a context-sensitive response based on a user generated input message and conversational context data. A digital assistant utilizes the trained response generation engine to generate context-sensitive, natural language responses that are pertinent to user queries.

Abstract: Systems, methods, and computer-executable instructions for approximating a softmax layer are disclosed. A small world graph that includes a plurality of nodes is constructed for a vocabulary of a natural language model. A context vector is transformed. The small world graph is searched using the transformed context vector to identify a top-K hypothesis. A distance from the context vector for each of the top-K hypothesis is determined. The distance is transformed to an original inner product space. A softmax distribution is computed for the softmax layer over the inner product space of the top-K hypothesis. The softmax layer is useful for determining a next word in a speech recognition or machine translation.

Abstract: A processing unit can acquire datasets from respective data sources, each having a respective unique data domain. The processing unit can determine values of a plurality of features based on the plurality of datasets. The processing unit can modify input-specific parameters or history parameters of a computational model based on the values of the features. In some examples, the processing unit can determine an estimated value of a target feature based at least in part on the modified computational model and values of one or more reference features. In some examples, the computational model can include neural networks for several input sets. An output layer of at least one of the neural networks can be connected to the respective hidden layer(s) of one or more other(s) of the neural networks. In some examples, the neural networks can be operated to provide transformed feature value(s) for respective times.

Abstract: Generally discussed herein are devices, systems, and methods for task routing. A method can include receiving, from a resource, a request for a task, in response to receiving the request, determining whether to retrieve a new task of new tasks stored in a first queue or a backlog task of backlog tasks stored in a second queue based on a combined amount of backlog tasks and new tasks relative to a capacity of the resource or the resources, retrieving the new task or the backlog task from the determined first queue or second queue, respectively, based on the determination, and providing the retrieved task to the resource.

Abstract: A system for executing composite tasks can include a processor to detect a composite task from a user. The processor can also detect a plurality of subtasks corresponding to the composite task based on unsupervised data without a label, wherein the plurality of subtasks are identified by a top-level dialog policy. The processor can also detect a plurality of actions, wherein each action is to complete one of the subtasks, and wherein each action is identified by a low-level dialog policy corresponding to the subtasks identified by the top-level dialog policy. The processor can also update a dialog manager based on a completion of each action corresponding to the subtasks and execute instructions based on a policy identified by the dialog manager, wherein the executed instructions implement the policy with a lowest global cost corresponding to the composite task provided by the user.

Abstract: A processing unit can successively operate layers of a multilayer computational graph (MCG) according to a forward computational order to determine a topic value associated with a document based at least in part on content values associated with the document. The processing unit can successively determine, according to a reverse computational order, layer-specific deviation values associated with the layers based at least in part on the topic value, the content values, and a characteristic value associated with the document. The processing unit can determine a model adjustment value based at least in part on the layer-specific deviation values. The processing unit can modify at least one parameter associated with the MCG based at least in part on the model adjustment value. The MCG can be operated to provide a result characteristic value associated with test content values of a test document.

Abstract: Systems and methods for determining knowledge-guided information for a recurrent neural networks (RNN) to guide the RNN in semantic tagging of an input phrase are presented. A knowledge encoding module of a Knowledge-Guided Structural Attention Process (K-SAP) receives an input phrase and, in conjunction with additional sub-components or cooperative components generates a knowledge-guided vector that is provided with the input phrase to the RNN for linguistic semantic tagging. Generating the knowledge-guided vector comprises at least parsing the input phrase and generating a corresponding hierarchical linguistic structure comprising one or more discrete sub-structures. The sub-structures may be encoded into vectors along with attention weighting identifying those sub-structures that have greater importance in determining the semantic meaning of the input phrase.