Abstract: A method of operating an electronic device and an electronic device thereof are provided. The method includes receiving a first voice signal of a first user, authenticating whether the first user has authority to control the electronic device, based on the first voice signal, and determining an instruction corresponding to the first voice signal based on an authentication result and controlling the electronic device according to the instruction. The electronic device includes a receiver configured to receive a first voice signal of a first user and at least one processor configured to authenticate whether the first user has authority to control the electronic device based on the first voice signal, determine an instruction corresponding to the first voice signal, and control the electronic device according to the instruction.

Abstract: There is provided an information processing device, an information processing terminal, and an information processing method which are capable of presenting a choice which is easily recognized by voice. The information processing device according to one aspect of the present technology acquires a plurality of pieces of presentation information to be presented as choices and causes a part which is not similar to the other presentation information among the respective pieces of presentation information to be presented in a form different from a similar part. The present technology can be applied to devices having a voice assistant function of assisting a behavior of a user.

Abstract: Disclosed herein is method and system for detecting intent as an ordered sequence from user query. The system identifies word embedding feature for each word using word embedding model and identifies Part of Speech tag feature. Thereafter, system determines dependency label feature based on dependency role and POS tag feature. Further, system provides feature vector comprising POS tag feature of target word, POS tag feature of previous two words of target word, word embedding feature of target word, word embedding feature of head word for each target word and dependency label feature of target word to deep neural network for detecting intent as ordered sequence. The ordered sequence includes desired action in user query, focal point pertaining to which desired action must be performed and one or more descriptors associated with focal point. In this manner, in present disclosure, overall intent from user query is captured for accurately providing response.

Abstract: A context-aware account management system is described. The system receives a partial request and parses the partial request to identify one or more characters in the partial request. The system determines a plurality of valid requests that include the one or more characters, for instance in response to querying one or more data structures and/or generating the valid requests using a trained machine learning model. The system generates a suggested request by selecting one of the plurality of valid requests, and in some cases, by identifying valid values for variables within the selected one of the plurality of valid requests by querying one or more data structures and/or generating the valid values using a trained machine learning model. The system outputs the suggested request and receives a confirmation in response. The system performs an action indicated by the suggested request using a user account indicated by the suggested request.

Abstract: A method for conducting an automated interview session between a candidate and an automated chat resource includes analyzing a candidate resume to identify a first personality trait; generating a first resume-based question based on the first personality trait; presenting, by the automated chat resource, the first resume-based question to the candidate during the automated interview session based on the identified first personality trait; generating a first trait score based on a first response received from the candidate, wherein the first trait score is indicative of a relevance of the identified first personality trait; generating a first question-answer score based on the first resume-based question and the first response, wherein the first question-answer score is indicative of a first question-answer relevance and is used to determine whether to present a second resume-based question to the candidate; analyzing a communication transcript to identify a second personality trait; generating a first transcr

Abstract: Identifying groupings of individual users of a mobile network. In particular, some embodiments are configured to obtain transaction data records from one or more mobile network provider(s). Groupings of users may be identified by capturing association patterns in the data by applying word embedding.

Abstract: Systems and methods are provided that that may be implemented to extend functionality of a voice assistant and/or telecommunications software to a remote location via an endpoint device.

Abstract: A method for performing domain adaptation of an end-to-end (E2E) automatic speech recognition (ASR) model. The method includes: obtaining an un-adapted version of the E2E ASR model trained using a first set of transcriptions, the un-adapted version of E2E ASR model including an encoder network, a first prediction network and a joint network; using the first set of transcriptions, while keeping parameters of first prediction network fixed, to train a language model output component to match the first prediction network; using a second set of transcriptions, while keeping parameters of language model output component fixed, to fine-tune the first prediction network for obtaining a second prediction network; and generating an adapted version of the E2E ASR model, wherein the adapted version of the E2E ASR model comprises the encoder network, the second prediction network, the language model output component, and the joint network.

Abstract: Provided is a system for generating speech recognition training data, the system including: a speech data processing module receiving speech data from a user terminal and performing data preprocessing on the received speech data; an auto speech recognition (ASR) interfacing module transmitting the preprocessed speech data to a plurality of ASR engines and acquiring a confidence score and transcription data of the speech data from the plurality of ASR engines; an ASR result evaluating module determining whether the speech data and the transcription data match each other; and a training data managing unit generating training data as a pair of the speech data and the transcription data determined to match each other.

Abstract: This disclosure relates to systems and methods for discovering intents from multimedia conversations. In an implementation, the system may obtain an optimal parameter configuration based on a topic relevance quantifier. The system may extract topics from the audio calls based on the optimal parameter configuration. The system may determine respective context metrics for each of the topics. The system may select a relevant topic with respect to the intent discovery from the extracted topic based on the respective context metrics. The system may obtain documents associated with the relevant topic from the corpus of documents. The system may display information of the documents associated with the relevant topic via a user interface. The system may receive a feedback data with respect to the relevant topic. The system may determine intent information associated with the relevant topic based on the feedback data.

Abstract: Embodiments are directed to organizing conversation information. A tracker vocabulary may be provided to a universal model to predict a generalized vocabulary associated with the tracker vocabulary. A tracker model may be generated based on the portions of the universal model activated by the tracker vocabulary such that a remainder of the universal model may be excluded from the tracker model. Portions of a conversation stream may be provided to the tracker model. A match score may be generated based on the track model and the portions of the conversation stream such that the match score predicts if the portions of the conversation stream may be in the generalized vocabulary predicted for the tracker vocabulary. Tracker metrics may be collected based on the portions of the conversation and the match scores such that the tracker metrics may be included in reports or notifications.

Abstract: The present disclosure discloses a control method, including: acquiring an adjustment instruction, acquiring a target device, and processing the adjustment instruction such that the target device responds to the adjustment instruction. The adjustment instruction is determined according to a first sound information and the target device is determined based on information that characterizes an ambient sound. A control device and a control system are also provided.

Abstract: Systems and methods for speech recognition are provided. The method may include obtaining a plurality of candidate recognition results of speech information uttered by a user and a plurality of preliminary scores corresponding to the plurality of candidate recognition results, respectively. The method may further include, for each of the plurality of candidate recognition results, extracting one or more keywords from the candidate recognition result and determining at least one parameter associated with the one or more extracted keywords. The method may further include, for each of the plurality of candidate recognition results, generating an updating coefficient based on the at least one parameter and updating the preliminary score based on the updating coefficient to generate an updated score. The method may further include determining, from the plurality of candidate recognition results, a target recognition result based on the plurality of updated scores.

Abstract: The present disclosure provides a voiceprint recognition method and apparatus, comprising: according to an obtained command speech, recognizing, in a voiceprint recognition manner, a user class sending a command speech; according to the user class, using a corresponding speech recognition model to perform speech recognition for the command speech, to obtain a command described by the command speech; providing resources according to the user class and command. The present disclosure can avoid the problems that in a conventional voiceprint recognition method in the prior art, a client needs to participate in voiceprint recognition, and the user's ID needs to be further recognized through a voiceprint training process, and that the user's degree of satisfaction is not high. While the user speaks naturally, it is feasible to perform processing for these very “ordinary” speech, and meanwhile complete the work of voiceprint recognition.

Abstract: A method of disambiguating user queries in a multi-turn dialogue include a set of user utterances includes using a predefined language model to recognize an ambiguous entity in an unresolved user utterance from the multi-turn dialogue. The method further includes outputting a clarifying question about the ambiguous entity, and receiving a clarifying user utterance. The method further includes identifying a disambiguating entity in the clarifying user utterance. The method further includes rewriting the unresolved user utterance as a rewritten utterance that replaces the ambiguous entity with the disambiguating entity, and outputting the rewritten utterance to one or more query answering machines.

Abstract: A processor-implemented response inference method and apparatus are disclosed. The response inference apparatus receives an input, generates a latent variable vector in a latent variable region space by encoding the input, generates a validation vector with a predetermined phase difference from the latent variable vector, generates an output response by decoding the latent variable vector, generates a validation response by decoding the validation vector, and validates the output response by comparing the output response to the validation response.

Abstract: A system and method are disclosed capable of parsing a spoken utterance into a natural language request and a speech audio segment, where the natural language request directs the system to use the speech audio segment as a new wakeword. In response to this wakeword assignment directive, the system and method are further capable of immediately building a new wakeword spotter to activate the device upon matching the new wakeword in the input audio. Different approaches to promptly building a new wakeword spotter are described. Variations of wakeword assignment directives can make the new wakeword public or private. They can also add the new wakeword to earlier wakewords, or replace earlier wakewords.

Abstract: A chatbot system receives utterances of a conversation. The chatbot system constructs a conversation knowledge graph comprising one or more dialogue segments that correspond to utterances of the conversation. The chatbot system identifies a dialogue segment in the conversation knowledge graph having a contextual uncertainty that is detected based on a context model. The chatbot system generates a clarifying question for the identified dialogue segment having the contextual uncertainty. The chatbot system receives a clarifying answer from a user interface of the computing device to the clarifying question. The chatbot system updates the context model and the conversation knowledge graph based on the clarifying question and the clarifying answer to resolve the contextual uncertainty of the identified dialogue segment.

Abstract: Described techniques select portions of an audio stream for transmission to a trained machine learning application, which generates response recommendations in real-time. This real-time response is facilitated by the system identifying, selecting and transmitting those portions of the audio stream likely to be most relevant to the conversation. Portions of an audio stream less likely to be relevant to the conversation are identified accordingly and not transmitted. The system may identify the relevant portions of an audio stream by detecting events in a contemporaneous event stream, use a trained machine learning model to identify events in an audio stream, or both.

Abstract: The systems and methods of the present disclosure generally relate to a data processing system that can identify and surface alternative requests when presented with ambiguous, unclear, or other requests to which a data processing system may not be able to respond. The data processing system can improve the efficiency of network transmissions to reduce network bandwidth usage and processor utilization by selecting alternative requests that are responsive to the intent of the original request.