Abstract: A method for waking up a robot includes: acquiring sight range information when a voice command issuer issues a voice command; if the sight range information of the voice command issuer when issuing the voice command is acquired, determining, based on the sight range information, whether the voice command issuer gazes the robot when the voice command is issued; and determining that the robot is called if the voice command issuer gazes the robot.

Abstract: An information processing device includes a processor. The processor is configured to: receive an input of a question; hold a response, when data required to output response content in response to the question is insufficient; and output, when insufficient data is collected while the response is being held, an announcement that the response is made and the response content.

Abstract: Systems, methods, and devices for outputting indications regarding voice-based interactions are described. A first speech-controlled device detects spoken audio corresponding to recipient information. The first device captures the audio and sends audio data corresponding to the captured audio to a server. The server determines a second speech-controlled device of the recipient and sends a signal to the recipient's second speech-controlled device representing a message is forthcoming. The recipient's second speech-controlled device outputs and indication representing a message is forthcoming.

Abstract: Methods, apparatus, systems, and computer-readable media are provided for storing incomplete dialog sessions between a user and an automated assistant in order that the dialog sessions can be completed in furtherance of certain actions. While interacting with an automated assistant, a user can become distracted and not complete the interaction to the point of the automated assistant performing some action. In response, the automated assistant can store the interaction as a dialog session. Subsequently, the user may express interest, directly or indirectly, in completing the dialog session, and the automated assistant can provide the user with a selectable element that, when selected, causes the dialog session to be reopened. The user can then continue the dialog session with the automated assistant in order that the originally intended action can be performed by the automated assistant.

Abstract: Embodiments may process search input for different users based on classifications of information and based on emotional content of search commands from the users. For example, a method may comprise receiving, at a computer system, speech data from a client device, the speech data representing a voice command from a user, obtaining, at the computer system, a plurality of items of content responsive to the voice command by searching for content, determining, at the computer system, at least one class related to the voice command, classifying, at the computer system, each obtained item of content into at least one class, identifying, at the computer system, at least one item of content classified into at least one class related to the voice command, and transmitting, at the computer system, the at least one identified item of content.

Abstract: A method of disambiguating user queries in a multi-turn dialogue including a set of user utterances. The method includes using a predefined language model to recognize an ambiguous entity in an unresolved user utterance from the multi-turn dialogue, and using the predefined language model to recognize entity constraints of the ambiguous entity. The method further includes, in a computer-accessible conversation history of the multi-turn dialogue, searching a set of previously-resolved entities for a candidate entity having entity properties with a highest confidence correspondence to the entity constraints of the ambiguous entity. The unresolved user utterance is rewritten as a rewritten utterance that replaces the ambiguous entity with the candidate entity. The rewritten utterance is output to one or more query answering machines.

Abstract: Provided are embodiments for a computer-implemented method for interacting with a user by an automated response system supporting topic switching and information collection. The computer-implemented method includes receiving a plurality of utterances from the user by the automated response system, and analyzing the utterances to form a first topic thread and an information collection objective. The computer-implemented method also includes utilizing an information collection user interface to gather data to support the information collection objective, and providing responses to the user after the gathered data related to the first topic thread. Also provided are embodiments for a system and computer program product for implementing the techniques described herein.

Abstract: A computer system conducts a communication session using a communication agent. Organizational information relating to a user is received. A communication session is conducted between a user and a communication agent, wherein the communication agent discusses one or more topics of an organization of the user. Organizational analytics are generated by applying natural language processing to user feedback to identify user issues pertaining to the one or more organizational topics wherein the organizational analytics are based on user sentiment toward the one or more organizational topics. The organizational analytics are reported to a third party. Embodiments of the present invention further include a method and program product for conducting a communication session using a communication agent in substantially the same manner described above.

Abstract: Methods, apparatus, systems, and computer-readable media are provided for storing incomplete dialog sessions between a user and an automated assistant in order that the dialog sessions can be completed in furtherance of certain actions. While interacting with an automated assistant, a user can become distracted and not complete the interaction to the point of the automated assistant performing some action. In response, the automated assistant can store the interaction as a dialog session. Subsequently, the user may express interest, directly or indirectly, in completing the dialog session, and the automated assistant can provide the user with a selectable element that, when selected, causes the dialog session to be reopened. The user can then continue the dialog session with the automated assistant in order that the originally intended action can be performed by the automated assistant.

Abstract: A method of providing real-time translation for video chat is provided. The method includes: continuously receiving first-language voice data and at least one second-language word from a first terminal; continuously displaying the at least one second-language word at the same time as reproduction of the voice data; acquiring a second-language translation of an ended sentence included in a voice recognition result for the voice data; and substituting at least one word, which corresponds to the ended sentence in the displayed at least one second-language word, with the acquired translation. The at least one second-language word corresponds to respective words included in the voice recognition result for the voice data.

Abstract: Low energy deep-learning networks for generating auditory features such as mel frequency cepstral coefficients in audio processing pipelines are provided. In various embodiments, a first neural network is trained to output auditory features such as mel-frequency cepstral coefficients, linear predictive coding coefficients, perceptual linear predictive coefficients, spectral coefficients, filter bank coefficients, and/or spectro-temporal receptive fields based on input audio samples. A second neural network is trained to output a classification based on input auditory features such as mel-frequency cepstral coefficients. An input audio sample is provided to the first neural network. Auditory features such as mel-frequency cepstral coefficients are received from the first neural network. The auditory features such as mel-frequency cepstral coefficients are provided to the second neural network. A classification of the input audio sample is received from the second neural network.

Abstract: A computer-implemented method is provided that includes receiving a search query and, responsive to the search query, providing one or more textual comments relevant to the search query. This includes tokenizing the search query and calculating a set of query term frequency metrics. A set of records relevant to the search query is then selected, from a persistent storage, based on determined similarities between the query term frequency metrics and frequency metrics determined for the records in the persistent storage. Textual comments within the selected records are associated with usefulness metrics. The textual comments relevant to the search query are selected by selecting those textual comments within the selected records that are associated with usefulness metrics that are within a pre-determined range, e.g., an inter-quartile range for a population of usefulness metrics.

Abstract: Disclosed are a speech recognition method and a speech recognition device, in which speech recognition is performed by executing an artificial intelligence (AI) algorithm and/or a machine learning algorithm provided therein. According to one embodiment, the speech recognition method includes buffering a spoken utterance, extracting a standby wake-up word corresponding to a preset wake-up word from the spoken utterance by comparing the buffered spoken utterance to the preset wake-up word, analyzing the role of the standby wake-up word in the spoken utterance, determining the speech intent in uttering the standby wake-up word by using results of analyzing the role of the standby wake-up word, and determining whether to execute a spoken sentence as a voice command in the spoken utterance and processing the spoken sentence accordingly.

Abstract: Mechanisms are provided to perform embedding of content of a natural language document. The mechanisms receive a document data object of an electronic document and analyze a structure of the electronic document to identify one or more structural document elements that have a relationship with the document data object. A dependency data structure is generated, representing the electronic document, where edges define relationships between document elements and at least one edge represents at least one relationship between the one or more structural document elements and the document data object. The mechanisms embed the document data object based on the at least one relationship to thereby represent the document data object as a vector data structure. The mechanisms perform natural language processing on the portion of natural language content based on the vector data structure. The one or more structural document elements are non-local non-contiguous with the document data object.

Abstract: A system and method for engaging in an automated dialog with a user. A processor retrieves a preset dialog flow that includes various blocks directing the dialog with the user. The processor provides a prompt to the user based on a current block of the dialog flow, receives an action from the user in response to the prompt, and retrieves a classification/decision tree corresponding to the dialog flow. The classification tree has a plurality of nodes mapped to the blocks of the dialog flow. Each of the nodes represents a user intent. The processor computes a probability for each of the nodes based on the action from the user. A particular one of the nodes is then selected based on the computed probabilities. A target block of the dialog flow is further identified based on the selected node, and a response is output in response to the identified target block.

Abstract: A method of checking a link in a body of text comprises receiving the text and detecting a link to an external source within the received text. At least a portion of the received text is selected for analysis and one or more important keywords within the selected portion of the received text are determined. Text is obtained from the external source by accessing the link. At least a portion of the obtained text is selected for analysis and one or more important keywords within the selected portion of the obtained text are determined. The more important keywords within the selected portion of the original received text are compared with the important keywords within the selected portion of the obtained text from the link, and an output is provided depending upon the result of the comparison of the one or more important keywords within the selected portion of the received text with the one or more important keywords within the selected portion of the obtained text.

Abstract: A system and method for transforming unstructured text into structured form is disclosed. The system and method include converting an input word sequence (e.g., sentence) into tagged output which can be then easily be converted into a structured format. The system may include a bidirectional recurrent neural network that can generate multiple labels of individual words or phrases. In some embodiments, a customized learning loss equation involving set similarity is used to generate the multiple labels.

Abstract: A method for generating an automatic speech recognition (ASR) model using unsupervised learning includes obtaining, by a device, text information. The method includes determining, by the device, a set of phoneme sequences associated with the text information. The method includes obtaining, by the device, speech waveform data. The method includes determining, by the device, a set of phoneme boundaries associated with the speech waveform data. The method includes generating, by the device, the ASR model using an output distribution matching (ODM) technique based on determining the set of phoneme sequences associated with the text information and based on determining the set of phoneme boundaries associated with the speech waveform data.

Abstract: An automated natural dialogue system provides a combination of structure and flexibility to allow for ease of annotation of dialogues as well as learning and expanding the capabilities of the dialogue system based on natural language interactions.

Abstract: Disclosed herein are system, method, and computer-readable storage-medium embodiments for automatic discovery of translated text. An embodiment may include relating a user-interface (UI) output with a corresponding localization object in a code-base index and matching a first instance of a unique identifier with a second instance of the unique identifier. The first instance of the unique identifier may be located in a code base corresponding to the code-base index, and the second instance of the unique identifier may correspond to the UI output. The code base may be structured to comprise the unique identifier in a given context. Further operations may include retrieving a reference to the corresponding localization object of the UI output in response to a determination that the UI output is incorrect in the given context, and outputting the reference to the corresponding localization object. The reference may be copied into a ticket of a tracking system.