Abstract: Methods, and systems for named entity recognition and disambiguation. One of the methods includes: receiving a document; extracting a plurality of mentions within the document; clustering the plurality of mentions to produce at least one cluster; identifying a plurality of candidate entities for the cluster; pairing each candidate entity and the cluster to create a plurality of candidate entity-cluster pairings; generating features for each candidate entity-cluster pairing; and selecting an entity, from the plurality of candidate entities, for the cluster based at least in part on features of a candidate entity pairing.

Abstract: The present disclosure provides a technical solution related to establishing a characterized chatbot with personality. On one hand, a corpus database matched with a specific character may be established, and a generated regular/conventional response messages may be converted into a characterized response messages with character's features during conversation, so that the conversation style of a chatbot may be characterized with a certain character's features. On the other hand, a chatbot may select a response message with a specific emotion by using an emotion conversation table corresponding to a personality when generating the response message against a content of a conversation input by a user, so that a chatbot may be characterized and have certain personality.

Abstract: According to one embodiment, a computing device may conduct an active telephone conversation and a proximity of the computing device to a second computing device may be determined to be within a predetermined distance. Moreover, a determination of a capability of the second computing device to conduct a telephone conversation may be made, and in response to such determinations, the active telephone conversation may be transferred from the computing device to the second computing device. In an example, when a user holding a telephone conversation moves near a smart speaker, the telephone conversation may be transferred to the smart speaker from the user's mobile device, should the user desire such a transfer. Moreover, as the user moves around, the telephone conversation may move with the user from device to device, enabling a seamless and intuitive telephone conversation experience for the user.

Abstract: Based on a candidate set of translations produced by a neural network based machine learning model, a mapping data structure such as a statistical phrase table is generated. The mapping data structure is analyzed to obtain a quality metric of the neural network based model. One or more operations are initiated based on the quality metric.

Abstract: The present disclosure proposes a solution to associate a device with a user by capturing a voice of a speaker by a microphone connected to the network device (e.g. a residential or home gateway), monitoring the IP traffic of the network device and detecting the device contributing to this IP traffic in order to establish a link between the speaker and his device(s) and associate the device with the speaker.

Abstract: Disclosed is a system for telephones by providing an improved and streamlined user experience and enhanced fail over mechanisms. A decentralized system managed through a web site which allows for continued operation even when the primary systems fail includes a mechanism for restoring the primary systems automatically when they become available again. Phones connect to two PBX systems at the same time, one local and one at a remote location. The two PBX systems synchronize configuration data and media files between them. The website can also be used to manage any number of systems allowing any size organization to manage every phone system in their organization from a single interface.

Abstract: Sentence pairs can be filtered using dual monolingual conditional cross-entropy filtering. The filtered sentence pairs can be used to train a model for performing a language processing task, such as machine translation.

Abstract: Embodiments for managing virtual assistants are described. Information associated with a user in an internet of things (IoT) device environment having a plurality of IoT devices is received. A request from the user is received. In response to the receiving of the request, a first portion of a response to the request is caused to be rendered utilizing a first of the plurality of IoT devices. Movement of the user within the IoT device environment is detected. In response to the detecting of the movement of the user, a second portion of the response to the request is caused to be rendered utilizing a second of the plurality of IoT devices based on said detected movement of the user and said received information about the user.

Abstract: A channel-based voice call method, including: accepting a matching request of a client, acquiring a matching result of the client from a matching pool, and returning the matching result to the client; and acquiring a call matching acceptance instruction of the client, allocating a temporary channel and locking the temporary channel, and sending a successfully accepted matching message, a temporary channel number and a channel password to the client, such that the client enters the temporary channel based on the successfully accepted matching message, the temporary channel number and the channel password to make a call.

Abstract: An expert system designed to automatically generate requirements receives a set of natural-language statements. The system extracts from each statement keywords, rules, conditions, and semantic meanings of the statement's verbs. Verbs are organized into groups of synonyms and each statement's verb groups are each associated with one or more of that statement's keywords. Each statement is categorized as having simple, complex, or compound atomicity, as a function of the numbers of that statement's keywords, rules and verbs, and is assigned a degree of orthogonality as a function of the numbers of that statement's keywords that are associated with each of that statement's groups. Each statement is divided into segments and the procedure is repeated upon each segment until all resulting segments have simple atomicity. The segments are forwarded to downstream computerized modules that use the segments as requirements for a planned project.

Abstract: An electronic device for providing contextual recommendations and a method therefor are provided. The electronic device includes a memory storing instructions, and at least one processor configured to execute the instructions to obtain text including characters, identify, based on a language model, a domain corresponding to the text, identify a reference of the characters based on the domain, and perform, based on the reference, an action related to the domain.

Abstract: Aspects of the present disclosure include monitoring a contact's communication device while a communication session is in a hold state to determine a sentiment of the contact. Natural language recognition and natural language processing are used to determine a sentiment score of the contact based on overheard communications while the communication session is in the hold state. Then, an action is performed based on the sentiment score of the contact.

Abstract: Systems, methods, and devices are disclosed for detecting an active speaker in a two-way conference. Real time audio in one or more sub band domains are analyzed according to an echo canceller model. Based on the analyzed real time audio, one or more audio metrics are determined from output from an acoustic echo cancellation linear filter. The one or more audio metrics are weighted based on a priority, and a speaker status is determined based on the weighted one or more audio metrics being analyzed according to an active speaker detection model. For an active speaker status, one or more residual echo or noise is removed from the real time audio based on the one or more audio metrics.

Abstract: Sentence pairs can be filtered using dual monolingual cross-entropy-delta filtering. The filtered sentence pairs can be used to train a model for performing a language processing task, such as machine translation.

Abstract: A call reminder functionality is provided by an arrangement that is operable with a network-enabled communication device in which a reminder to return a phone call can be generated in an automated manner upon the occurrence of a call event. When the call event occurs, an option is displayed on a user interface (“UI”) on the device to create a link between the call and the user's schedule application. If the user selects the option, a callback task is added to the user's task list maintained by the schedule application, and a reminder to make the callback is generated and then displayed on the UI at a future time. The reminder includes pertinent call information and also includes a button that enables the user to place the callback through the UI (i.e., a “one-click call back”).

Abstract: Systems and methods for using neuromuscular information to improve speech recognition. The system includes a plurality of neuromuscular sensors arranged on one or more wearable devices and configured to continuously record a plurality of neuromuscular signals from a user, at least one storage device configured to store one or more trained statistical models for determining text based on audio input and the plurality of neuromuscular signals, at least one input interface configured to receive the audio input, and at least one computer processor programmed to obtain the audio input and the plurality of neuromuscular signals, provide as input to the one or more trained statistical models, the audio input and the plurality of neuromuscular signals or signals derived from the plurality of neuromuscular signals, and determine based, at least in part, on an output of the one or more trained statistical models, the text.

Abstract: Aspects of the disclosure relate to automated redistribution of queries to underutilized channels. A computing platform may monitor user traffic for one or more customer service communication channels. Subsequently, the computing platform may identify estimated wait times for a plurality of users to be served via the one or more channels. Then, the computing platform may initiate, via an intelligent virtual assistant, a communication with a given user of the plurality of users. Then, the computing platform may receive, via the intelligent virtual assistant, one or more attributes of a query of the given user. Then, the computing platform may select a channel of the one or more channels. Then, the computing platform may provide, to an enterprise agent associated with the selected channel, the one or more attributes of the query. Subsequently, the computing platform may direct the given user to the selected channel.

Abstract: Systems and methods for controlling hearing devices are disclosed. Each hearing device is worn by a user, the hearing devices being communicatively interconnected with each other for data exchange to form a hearing device system. A method includes acquiring a sound signal of each user with a microphone of the hearing device of the user; determining from the sound signal a speaker signal, which indicates, whether the user is a speaking user; determining a priority score for each user from the speaker signal of the user; outputting sound signals from speaking users to the users via the hearing devices of the users, such that sound signals from speaking users with a priority score smaller than a threshold are muted.

Abstract: Methods of operating headphones may involve filtering an input signal into a first filtered input signal and a second filtered input signal utilizing a filter. The second filtered input signal may be sent directly to a tactile vibration driver and tactile vibrations may be produced. A fixed, predetermined inverse transfer function may be applied to the first filtered input signal, generating an anti-wave signal. The anti-wave signal may be summed with the first filtered input signal, generating an output signal. Alternatively, a fixed, predetermined transfer function may be applied to the first filtered input signal, generating a modified input signal. The modified input signal may be subtracted from the first filtered input signal, generating an output signal. Audio sound waves may be produced with an acoustic driver responsive to the output signal, reducing effects of incidental acoustic noise generated by the tactile vibration driver.

Abstract: A computer system is provided. The computer system includes a memory and at least one processor configured to recognize one or more intent keywords in text provided by a user; identify an intent of the user based on the recognized intent keywords; select a workflow context based on the identified intent; determine an action request based on analysis of the text in association with the workflow context, wherein the action request comprises one or more action steps and the action steps comprise one or more data points; obtaining a workspace context associated with the user; and evaluate the data points based on the workspace context.