Abstract: Techniques for limiting natural language processing performed on input data are described. A system receives input data from a device. The input data corresponds to a command to be executed by the system. The system determines applications likely configured to execute the command. The system performs named entity recognition and intent classification with respect to only the applications likely configured to execute the command.

Abstract: A system capable of performing natural language understanding (NLU) using different application domains in parallel. A model takes incoming query text and determines a list of potential supplemental intent categories corresponding to the text. Supplemental applications within those categories are then identified as likely candidates for responding to the query. Application specific domains, including NLU components for the particular supplemental applications, are then activated and process the query text in parallel. Further, certain system default domains may also process incoming queries substantially in parallel with the supplemental applications. The different results are scored and ranked to determine highest scoring NLU results.

Abstract: A system and method is disclosed for predicting user behavior in response to various tasks and or/applications. This system can be a neural network-based joint model. The neural network can include a base neural network portion and one or more task-specific neural network portions. The artificial neural network can be initialized and trained using data from multiple users for multiple tasks and/or applications. This user data can be related to characteristics and behavior, including age, gender, geographic location, purchases, past search history, and customer reviews. Additional task-specific neural network portions can be added to the neural network and may be trained using a task-specific subset of the training data. The joint model can be used to predict user behavior in response to an identified task and/or application. The tasks and/or applications can relate to use of a website by users.

Abstract: Systems, methods, and devices for computer-generating responses and sending responses to communications when the recipient of the communication is unavailable are disclosed. An individual may send a message (either audio or text) to a recipient. The recipient may be unavailable to contemporaneously respond to the message (e.g., the recipient may be performing an action that makes is difficult or impractical for the recipient to contemporaneously respond to the audio message). When the recipient is unavailable, a response to the message is generated and sent without receiving an instruction from the recipient to do so. The response may be sent to the message originating individual, and content of the response may thereafter be sent to the recipient to receive feedback regarding the correctness of the response. Alternatively, the response content may first be sent to the recipient to receive the feedback, and thereafter the response may be sent to the message originating individual.

Abstract: Features are disclosed for generating predictive personal natural language processing models based on user-specific profile information. The predictive personal models can provide broader coverage of the various terms, named entities, and/or intents of an utterance by the user than a personal model, while providing better accuracy than a general model. Profile information may be obtained from various data sources. Predictions regarding the content or subject of future user utterances may be made from the profile information. Predictive personal models may be generated based on the predictions. Future user utterances may be processed using the predictive personal models.

Abstract: The disclosed technology relates to a distributed policy store. A system is configured to locate, in an index, an entry for a network entity, determine, based on the entry, a file identifier for a file containing a record for the network entity and an offset indicating a location of the record in the file. The system is further configured to locate the file in a distributed file system using the file identifier, locate the record in the file using the offset, and retrieve the record.

Abstract: The disclosed technology relates to assigning network agents to communication modules. A network policy system is configured to assign network agents to buckets based on an agent identifier of each agent. The network policy system can assign buckets to communication modules. When a failed communication module is detected, the network policy system can reassigning buckets assigned to the failed communication module to operational communication modules.

Abstract: The disclosed technology relates to a distributed policy store. A system is configured to locate, in an index, an entry for a network entity, determine, based on the entry, a file identifier for a file containing a record for the network entity and an offset indicating a location of the record in the file. The system is further configured to locate the file in a distributed file system using the file identifier, locate the record in the file using the offset, and retrieve the record.

Abstract: The disclosed technology relates to a network agent for generating platform specific network policies. A network agent is configured to receive a platform independent network policy from a network policy system, determine implementation characteristics of the network entity, generate platform specific policies from the platform independent network policy based on the implementation characteristics of the network entity, and implement the platform specific policies on the network entity.

Abstract: The disclosed technology relates to intent driven network management. A system is configured to maintain an inventory store comprising records for a set of network entities in a network, wherein each network entity in the set of network entities is associated with a record in the inventory store. The system receives a user intent statement comprising an action and a flow filter representing network data flows on which the action is to be applied and queries, based on the flow filter, the inventory store to identify a plurality of network entities in the set of network entities to which the user intent statement applies. The system generates a plurality of network policies that implement the user intent statement based on the plurality of network entities and the action and enforces the plurality network policies.

Abstract: The disclosed technology relates to a network agent for reporting to a network policy system. A network agent includes an agent enforcer and an agent controller. The agent enforcer is configured to implementing network policies on the system, access data associated with the implementation of the network policies on the system, and transmit, via an interprocess communication, the data to the agent controller. The agent controller is configured to generate a report including the data and transmit the report to a network policy system.

Abstract: Features are disclosed for generating predictive personal natural language processing models based on user-specific profile information. The predictive personal models can provide broader coverage of the various terms, named entities, and/or intents of an utterance by the user than a personal model, while providing better accuracy than a general model. Profile information may be obtained from various data sources. Predictions regarding the content or subject of future user utterances may be made from the profile information. Predictive personal models may be generated based on the predictions. Future user utterances may be processed using the predictive personal models.

Abstract: An automatic speech recognition (“ASR”) system produces, for particular users, customized speech recognition results by using data regarding prior interactions of the users with the system. A portion of the ASR system (e.g., a neural-network-based language model) can be trained to produce an encoded representation of a user's interactions with the system based on, e.g., transcriptions of prior utterances made by the user. This user-specific encoded representation of interaction history is then used by the language model to customize ASR processing for the user.

Abstract: Features are disclosed for modeling user interaction with a detection system using a stochastic dynamical model in order to determine or adjust detection thresholds. The model may incorporate numerous features, such as the probability of false rejection and false acceptance of a user utterance and the cost associated with each potential action. The model may determine or adjust detection thresholds so as to minimize the occurrence of false acceptances and false rejections while preserving other desirable characteristics. The model may further incorporate background and speaker statistics. Adjustments to the model or other operation parameters can be implemented based on the model, user statistics, and/or additional data.

Abstract: Features are disclosed for detecting words in audio using contextual information in addition to automatic speech recognition results. A detection model can be generated and used to determine whether a particular word, such as a keyword or “wake word,” has been uttered. The detection model can operate on features derived from an audio signal, contextual information associated with generation of the audio signal, and the like. In some embodiments, the detection model can be customized for particular users or groups of users based usage patterns associated with the users.

Abstract: A two-way speech-to-speech (S2S) translation system actively detects a wide variety of common error types and resolves them through user-friendly dialog with the user(s). Examples include features including one or more of detecting out-of-vocabulary (OOV) named entities and terms, sensing ambiguities, homophones, idioms, ill-formed input, etc. and interactive strategies for recovering from such errors. In some examples, different error types are prioritized and systems implementing the approach can include an extensible architecture for implementing these decisions.

Abstract: Features are disclosed for filtering portions of an output audio signal in order to improve automatic speech recognition on an input signal which may include a representation of the output signal. A signal that includes audio content can be received, and a frequency or band of frequencies can be selected to be filtered from the signal. The frequency band may correspond to a desired frequency band for speech recognition. An input signal can be obtained comprising audio data corresponding to a user utterance and presentation of the output signal. Automatic speech recognition can be performed on the input signal. In some cases, an acoustic model trained for use with such frequency band filtering may be used to perform speech recognition.

Abstract: Features are disclosed for detecting words in audio using environmental information and/or contextual information in addition to acoustic features associated with the words to be detected. A detection model can be generated and used to determine whether a particular word, such as a keyword or “wake word,” has been uttered. The detection model can operate on features derived from an audio signal, contextual information associated with generation of the audio signal, and the like. In some embodiments, the detection model can be customized for particular users or groups of users based usage patterns associated with the users.

Abstract: Features are disclosed for estimating a false rejection rate in a detection system. The false rejection rate can be estimated by fitting a model to a distribution of detection confidence scores. An estimated false rejection rate can then be computed for confidence scores that fall below a threshold. The false rejection rate and model can be verified once the detection system has been deployed by obtaining additional data with confidence scores falling below the threshold. Adjustments to the model or other operational parameters can be implemented based on the verified false rejection rate, model, or additional data.

Abstract: Techniques are described for documenting the positions of items in a room, such as in rooms that are configured for testing automated systems that perform position-related functions. A non-contact measuring tool may be placed at different reference positions within the room. At each position, measurements are made to the room corners and to items of interest within the room. Based on this information, coordinates of the reference positions are calculated. Coordinates of the items are calculated based on the determined coordinates of the reference positions.