Microsoft Technology Licensing, LLC

Abedelkader Asi

Royi Ronen

Erez Altus

Yarin Kuper

Weizhu Chen

Ohad Jassin

Roy Eisenstadt

Yi Mao

Ran Mizrachi

Dean Geckt

Song Wang

Rona Shaanan

Olga Goldenberg

Liron Izhaki-Allerhand

Tomer Rosenthal

Shimrit Rada Bemis

The disclosure herein describes reducing training bias in outputs generated by a generative language model. A communication segment associated with a communication is obtained by at least one processor of a generative language model. An output value associated with the communication segment is generated by the generative language model. The output value is mapped to a set of training bias values associated with the generative language model and based on the mapping of the output value to a training bias value of the set of training bias values, an alternative output value is generated. The alternative output value is used in a generated segment output for the communication segment. The accuracy of segment outputs generated by the generative language model is improved through reducing or eliminating its training biases.

Reducing biases of generative language models

A method of generating keyword-based dialogue summaries is provided. The method includes inputting a transcript of an audio conversation and a keyword into a machine learning model trained based on encodings representing the keyword and the transcript, generating computer-generated text different from and semantically descriptive of the transcript and semantically associated with the keyword, and outputting the computer-generated text in association with a selectable item selectable for inclusion of the computer-generated text in displayed text representing the transcript, the selectable item associated with the keyword.

Keyword-based dialogue summarizer

The disclosure herein describes determining topics of communication transcripts using trained summarization models. A first communication transcript associated with a first communication is obtained and divided into a first set of communication segments. A first set of topic descriptions is generated based on the first set of communication segments by analyzing each communication segment of the first set of communication segments with a generative language model. A summarization model is trained using the first set of communication segments and associated first set of topic descriptions as training data. The trained summarization model is then applied to a second communication transcript and, based on applying the trained summarization model to the second communication transcript, a second set of topic descriptions of the second communication transcript is generated. By training the summarization model based on output of the generative language model, it enables efficient, accurate generation of topic descriptions from communication transcripts.

Determining topic labels for communication transcripts based on a trained generative summarization model

Technology is disclosed for providing dynamic identification and extraction or tagging of contextually-coherent text blocks from an electronic document. In an embodiment, an electronic document may be parsed into a plurality of content tokens that each corresponds to a portion of the electronic document, such as a sentence or a paragraph. Employing a sliding window approach, a number of token groups are independently analyzed, where each group of tokens has a different number of tokens included therein. Each token group is analyzed to determine confidence scores for various determinable contexts based on content included in the token set. The confidence scores can then be processed for each token group to determine an entropy score for the token group. In this way, one of the analyzed token groups can be selected as a representative text block that corresponds to one of the plurality of determinable contexts. A corresponding portion of the electronic document can be tagged with a corresponding context determined based on the analyzed content included therein, and provided for output.

Dynamic extraction of contextually-coherent text blocks

Growing community of inventors

Kfar Bara, Israel

Abedelkader Asi