Abstract: This is a system for sentiment analysis and misinformation detection for text transcriptions of conversations and statement. More particularly, this is a software system that uses neural network, machine learning (ML) models, which have been trained solely on text sourced from the internet, for the classification of the sentiment and risk of misinformation conveyed through a piece of text as it relates to a specified subject matter. Such a system typically is used on digital communication platforms, such as social media sites, to enable the determination of public opinion and perception in order to provide competitive intelligence for decision-making.

Abstract: Examples of the present disclosure describe systems and methods for intelligent contract analysis and data organization. In example aspects, input data may be received into the system in the form of a legal document, such as a contract. The contract may be processed by a natural-language processor, and a summarized version of the contract and/or particular contract provisions may be provided. In other aspects, a machine-learning engine may process the input data in combination with a user's decisions to accept or reject particular contract provisions. The ML engine may provide an intelligent recommendation to a user regarding whether the user should consent or not consent to a particular contract provision. In other example aspects, a data dashboard displaying how a user's personally identifiable information (PII) is shared with third parties may be displayed. Aggregated PII may further be displayed in a dashboard.

Abstract: The present disclosure is directed to systems and methods that include and/or leverage one or more machine-learned language models that generate intermediate textual analysis (e.g., including usage of structural tools such as APIs) in service of contextual text generation. For example, a computing system can obtain a contextual text string that includes one or more contextual text tokens. The computing system can process the contextual text string with the machine-learned language model to generate one or more intermediate text strings that include one or more intermediate text tokens. The computing system can process the one or more intermediate text strings with the machine-learned language model to generate an output text string comprising one or more output text tokens. The one or more intermediate text strings can include textual analysis of the contextual text string that supports the output text string.

Abstract: A decoder includes a feature extraction circuit for calculating one or more feature vectors. An acoustic model circuit is coupled to receive one or more feature vectors from and assign one or more likelihood values to the one or more feature vectors. A memory architecture that utilizes on-chip state lattices and an off-chip memory for storing states of transition of the decoder is used to reduce reading and writing to the off-chip memory. The on-chip state lattice is populated with at least one of the states of transition stored in the off-chip memory. An on-chip word is generated from a snapshot from the on-chip state lattice. The on-chip state lattice and the on-chip word lattice act as an on-chip cache to reduce reading and writing to the off-chip memory.

Abstract: The disclosure relates to an electronic apparatus for recognizing user voice and a method of recognizing, by the electronic apparatus, the user voice. According to an embodiment, the method of recognizing the user voice includes obtaining an audio signal segmented into a plurality of frame units, determining an energy component for each filter bank by applying a filter bank distributed according to a preset scale to a frequency spectrum of the audio signal segmented into the frame units, smoothing the determined energy component for each filter bank, extracting a feature vector of the audio signal based on the smoothed energy component for each filter bank, and recognizing the user voice in the audio signal by inputting the extracted feature vector to a voice recognition model.

Abstract: An embodiment extracts a set of designated entities and a set of relationships between designated entities from speech content of an audio feed of a plurality of participants of a current web conference using a machine learning model trained to classify parts of speech content. The embodiment generates a list of current action items based on the extracted set of designated entities and relationships between designated entities. The embodiment identifies a first current action item that is an updated version of an ongoing action item on a progress list of ongoing action items from past web conferences. The embodiment also identifies a second current action item that is unrelated to any of the ongoing action items on the progress list. The embodiment updates the progress list to include updates for the first current action item and by adding the second current action item.

Abstract: An electronic device may include: a memory; a sound sensor; and a processor, wherein the processor is configured to: receive, from the sound sensor, sound data including a first piece of data corresponding to a first frequency band and a second piece of data corresponding to a second frequency band different from the first frequency band; receive voice data related to a voice of a registered user from the memory; perform voice identification by comparing the first piece of data and the second piece of data with the voice data related to the voice of the registered user; and determine an output based on a result of the voice identification.

Abstract: A facility for identifying multi-word expressions in a natural language sentence is described. The facility provides the sentence to each of multiple natural language processing modules including a first module, a second module, and a third module. Each natural language processing module uses a different approach to identify a multi-word expression and a type of the multi-word expression. Upon determining that the multiple identifiers of the multi-word expression differ, the facility determines the multi-word expression using a resolution process, which can involve a logical rule set or a machine learning model.

Abstract: Disclosed herein are example techniques to support multiple voice assistant services. An example implementation may involve a playback device capturing audio from the one or more microphones into one or more buffers as a sound data stream monitoring the sound data stream for a wake word associated with a specific voice assistant service and monitoring the sound data stream for a wake word associated with the media playback system. The playback device generates a second wake-word event corresponding to a voice input when sound data matching the wake word associated with the media playback system in a portion of the sound data stream is detected. The playback device determines that the voice input includes sound data matching one or more playback commands and sends sound data representing the voice input to a voice assistant associated with the media playback system for processing of the second voice input.

Abstract: A method may include obtaining first audio data originating at a first device during a communication session between the first device and a second device. The method may also include obtaining an availability of revoiced transcription units in a transcription system and in response to establishment of the communication session, selecting, based on the availability of revoiced transcription units, a revoiced transcription unit instead of a non-revoiced transcription unit to generate a transcript of the first audio data. The method may also include obtaining revoiced audio generated by a revoicing of the first audio data by a captioning assistant and generating a transcription of the revoiced audio using an automatic speech recognition system. The method may further include in response to selecting the revoiced transcription unit, directing the transcription of the revoiced audio to the second device as the transcript of the first audio data.

Abstract: According to an embodiment, an electronic device comprises a communication module, a memory, and a processor configured to, upon obtaining a first utterance related to a memory service, prepare to store first information for the memory service for the first utterance and store the first information including essential information, sensitivity information for the first information, and an authentication method for the first information, detected from the first utterance, in the memory, and obtain a second utterance for looking up information related to the memory service, upon identifying that the obtained second utterance is one for looking up the first information, complete authentication based on the authentication method, and provide the essential information by a providing method determined based on the sensitivity information. Various other embodiments may be provided.

Abstract: Implementations set forth herein relate to an automated assistant that can be customized by a user to provide custom assistant responses to certain assistant queries, which may originate from other users. The user can establish certain custom assistant responses by providing an assistant response request to the automated assistant and/or responding to a request from the automated assistant to establish a particular custom assistant response. In some instances, a user can elect to establish a custom assistant response when the user determines or acknowledges that certain common queries are being submitted to the automated assistant—but the automated assistant is unable to resolve the common query. Establishing such custom assistant responses can therefore condense interactions between other users and the automated assistant.

Abstract: According an embodiment, an information processing device includes a memory and one or more processors coupled to the memory. The one or more processors are configured to: cluster key phrases extracted from a document group to generate clusters each including items; extract a sub-document group corresponding to a specified key phrase from the document group; calculate a score for each of the clusters, the score indicating usefulness of information displayed in a graph image that indicates an amount of information for each of the items in the sub-document group; select clusters to be displayed whose scores are from a highest ranking to a predetermined ranking, among the clusters; and generate the graph image of each of the clusters to be displayed.

Abstract: A question answering system includes: a first encoder module configured to receive a question, the question including a first plurality of words, and encode the question into a first vector representation; a second encoder module configured to encode a document into a second vector representation, the document including a second plurality of words; a first reading module configured to generate a third vector representation based on the first and second vector representations; a first reformulation module configured to generate a first reformulated vector representation based on the first vector representation; a second reading module configured to generate a fifth vector representation based on the second vector representation and the first reformulated vector representation; a second reformulation module configured to generate a second reformulated vector representation based on first reformulated vector representation; and an answer module configured to determine an answer to the question based on the secon

Abstract: A speech signal is received by a device comprising first and second transducers, and the first transducer comprises a microphone. A method comprises performing a first voice biometric process on speech contained in a first part of a signal received by the microphone, in order to determine whether the speech is the speech of an enrolled user. A first correlation is determined, between said first part of the signal received by the microphone and a corresponding part of the signal received by the second transducer. A second correlation is determined, between said second part of the signal received by the microphone and the corresponding part of the signal received by the second transducer. It is then determined whether the first correlation and the second correlation satisfy a predetermined condition.

Abstract: Information indicating what technology and service are necessary to solve issues is provided by a device including dictionaries recording words having similar meanings and concepts concept co-occurrences in which two of the concepts are combined, a classification of each of technology names and service names, and a score assigned to each of the classifications, a sentence input unit receiving a sentence representing an issue, a morpheme analysis unit performing morpheme analysis on the sentence and output morphemes, a concept conversion unit extracting a word from the morphemes and convert the extracted word to a concept, a classification identifying unit configured to sum scores of respective classifications in which a set of the extracted concepts and the concept co-occurrence match and identify the classification having a highest summed score, and a recommendation unit configured to obtain, from the name list, the technology names and service names belonging to the identified classification.

Abstract: Techniques are provided for using context tags in named-entity recognition (NER) models. In one particular aspect, a method is provided that includes receiving an utterance, generating embeddings for words of the utterance, generating a regular expression and gazetteer feature vector for the utterance, generating a context tag distribution feature vector for the utterance, concatenating or interpolating the embeddings with the regular expression and gazetteer feature vector and the context tag distribution feature vector to generate a set of feature vectors, generating an encoded form of the utterance based on the set of feature vectors, generating log-probabilities based on the encoded form of the utterance, and identifying one or more constraints for the utterance.

Abstract: Methods of authenticating a user or speaker are provided. These methods include obtaining an input speech signal and user credentials identifying the user or speaker. The input speech signal includes a single-channel signal or a multi-channel speech signal. The methods further include extracting a speech voiceprint from the input speech signal, and retrieving a reference voiceprint associated to the user credentials. The methods still further include determining a voiceprint correspondence between the speech voiceprint and the reference voiceprint, and authenticating the user or speaker depending on said voiceprint correspondence. The methods yet further include updating the reference voiceprint depending on the speech voiceprint corresponding to the authenticated user or speaker. Computer programs, systems and computing systems are also provided which are suitable for performing said methods of authenticating a user or speaker.

Abstract: Systems and processes for operating a digital assistant are provided. In an example process, low-latency operation of a digital assistant is provided. In this example, natural language processing, task flow processing, dialogue flow processing, speech synthesis, or any combination thereof can be at least partially performed while awaiting detection of a speech end-point condition. Upon detection of a speech end-point condition, results obtained from performing the operations can be presented to the user. In another example, robust operation of a digital assistant is provided. In this example, task flow processing by the digital assistant can include selecting a candidate task flow from a plurality of candidate task flows based on determined task flow scores. The task flow scores can be based on speech recognition confidence scores, intent confidence scores, flow parameter scores, or any combination thereof. The selected candidate task flow is executed and corresponding results presented to the user.

Abstract: Audio signals of speech may be processed using an acoustic model. An acoustic model may be implemented with multiple streams of processing where different streams perform processing using different dilation rates. For example, a first stream may process features of the audio signal with one or more convolutional neural network layers having a first dilation rate, and a second stream may process features of the audio signal with one or more convolutional neural network layers having a second dilation rate. Each stream may compute a stream vector, and the stream vectors may be combined to a vector of speech unit scores, where the vector of speech unit scores provides information about the acoustic content of the audio signal. The vector of speech unit scores may be used for any appropriate application of speech, such as automatic speech recognition.