International Business Machines Corporation

Microsoft Technology Licensing, LLC

Amir Kantor

David Amid

David Boaz

Luis Alfonso Lastras-Montano

Tin Kam Ho

Neil Rohit Mallinar

Guy Uziel

Shiraz Jitendra Cupala

Ateret Anaby-Tavor

Segev Shlomov

Ateret Anaby Tavor

Kevin Michael Carter

Shai Ifrach

Shlomi Maliah

Yasemin Cenberoglu

Kumud Ranjan

Shyam Gupta

Gautam Jain

Xiao Yan Lu

Ido Priness

Jaime Brooks Teevan

Boaz Carmeli

Adi L Miller

Michael Masin

Lan Ye

Oren Sar-Shalom

Naama Tepper

George Kour

Naama Zwerdling

Esther Goldbraich

Matthew Jonathan Gardner

Rotem Dror

Eleonora Shtotland

Gilad Gildin

Ofek David

A system for providing a personalized assistant within a network-based communication service includes one or more processors and a memory storage device storing instructions thereon. During a network-based communication session, the system receives a query from a computing device of a first communication session participant and processes the query by determining that a second communication session participant has shared content via a content sharing feature of the network-based communication service. In response, the system provides the query and at least a portion of the shared content as input to a model, which dynamically constructs a prompt for use as input with a generative language model. The system then presents the prompt as input to the generative language model, receives a response as output from the generative language model, and causes presentation of the response to be presented to the communication session participant.

Network-based communication session copilot

A system for providing a personalized assistant within a network-based communication session includes a processor and a memory storage device storing instructions. The system determines when a first communication session participant joins the network-based communication session after a threshold duration of time subsequent to the start time of the session. Upon determining the first participant has joined, the system obtains content associated with the session and creates request data for a pre-trained generative language model. The request data includes an instruction requesting a predetermined number of suggested utterances not present in the content, each utterance relating to one or more topics corresponding to the content. The system transforms the request data to a command based on a command template and provides the command to the generative language model. The system receives a response from the model, including the predetermined number of suggested utterances, and presents them to the communication session participant in a graphical user interface while the session is in session.

A method comprising using at least one hardware processor for receiving sensory data from at least one physical or virtual sensor. The hardware processor(s) are used for computing a plurality of decision options for configuration of the at least one physical or virtual sensor. The hardware processor(s) are used for computing a plurality of utility functions, and for each utility function: (a) computing a utility value for each decision option, and (b) identifying a first subset of decision options that substantially maximize the computed utility values. The hardware processor(s) are used for selecting at least one cross-function decision option from of the first subsets, wherein the at least one cross-function decision option is included in a substantially maximum number of the first subsets. The hardware processor(s) are used for applying at least one of the at least one cross-function decision options, to at least one physical or virtual sensor.

Optimizing combined sensory data by utilitarian selection of a subgroup of available sensors

Embodiments of the present systems and methods may provide techniques for augmenting textual data that may be used for textual classification tasks. Embodiments of such techniques may provide the capability to synthesize labeled data to improve text classification tasks. Embodiments may be specifically useful when only a small amount of data is available, and provide improved performance in such cases. For example, in an embodiment, a method implemented in a computer system may comprise a processor, memory accessible by the processor, and computer program instructions stored in the memory and executable by the processor, and the method may comprise fine-tuning a language model using a training dataset, synthesizing a plurality of samples using the fine-tuned language model, filtering the plurality of synthesized samples, and generating an augmented training dataset comprising the training dataset and the filtered plurality of synthesized sentences.

Language-model-based data augmentation method for textual classification tasks with little data

A method, apparatus, and system are provided for resolving conflicts between training data conflicts by retrieving independent training data sets, each comprising a plurality of intents and end-user utterances for use in training one or more classifiers to recognize a corresponding intent from one or more of the end-user utterances, providing a first test end-user utterance associated with a first intent from the first independent training data set to the one or more classifiers to select an output intent generated by the one or more classifiers; identifying a first conflict when the first intent does not match the output intent, and automatically generating, by the system, one or more conflict resolution recommendations for display and selection by an end user to resolve the first conflict.

Method, apparatus, and system for conflict detection and resolution for competing intent classifiers in modular conversation system

An example system includes a processor to receive data for a multi-objective task. The processor is to also perform the multi-objective task on the received data via a trained primal network. The primal network and a dual network are trained for a multi-objective task using a Lagrangian loss function representing a number of objectives. The primal network is trained to minimize the Lagrangian loss function and the dual network is trained to maximize the Lagrangian loss function.

Performing multi-objective tasks via primal networks trained with dual networks

An example system includes a processor to receive a prefix of conversation and a text input. The processor is to also generate a completed response based on the prefix of conversation and the text input via a trained primal network. The primal network is trained to minimize a Lagrangian loss function representing a number of objectives and a dual network is trained to maximize the Lagrangian loss function.

Generating completed responses via primal networks trained with dual networks

A method, computer system, and a computer program product for enhanced rating predictions is provided. The present invention may include receiving a user input. The present invention may then include translating the received user input into an embedding matrix and inputting the embedding matrix into a deep neural network. The present invention may further include generating, by the deep neural network, an output vector, a user bias term and an item bias term based on the embedding matrix. The present invention may then include calculating a predicted rating based on the generated output vector, the generated user bias term and the generated item bias term. The present invention may then include determining an accuracy of the predicted rating.

Enhancing rating prediction using reviews

A method comprising operating a computerized chatbot to: calculate first and second scores representing a relevance of input received from a user to functionalities provided by respective first and second modules, respectively, of the chatbot; associate the first and second modules with respective first and second names; introducing the modules to the user using their associated names; selecting a module to interact with the user based on at least one of: a name mentioned by the user and a score and switching between the first and second modules based on the names.

System and method for operating a chatbot

A method, system, and apparatus are provided for resolving conflicts between training data conflicts by retrieving independent training data sets, each comprising a plurality of intents and end-user utterances for use in training one or more classifiers to recognize a corresponding intent from one or more of the end-user utterances, providing a first test end-user utterance associated with a first intent from the first independent training data set to the one or more classifiers to select an output intent generated by the one or more classifiers; identifying a first conflict when the first intent does not match the output intent, and automatically generating, by the system, one or more conflict resolution recommendations for display and selection by an end user to resolve the first conflict.

Growing community of inventors

Haifa, Israel

Amir Kantor