Patents by Inventor Sandeep SUBRAMANIAN
Sandeep SUBRAMANIAN has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Publication number: 20230394308Abstract: There is provided a non-transitory storage medium and a system for generating an abstractive summary of a document using an abstractive machine learning algorithm (MLA). A document including a plurality of text sequences is received. An extractive summary of the document is generated, the extractive summary including a set of summary text sequences which is a subset of the plurality of text sequences. The abstractive MLA generates, based on the set of summary text sequences and at least a portion of the plurality of text sequences, an abstractive summary of the document including a set of abstractive text sequences, at least one abstractive text sequence not being included in the plurality of text sequences.Type: ApplicationFiled: July 19, 2023Publication date: December 7, 2023Applicant: ServiceNow Canada Inc.Inventors: Sandeep SUBRAMANIAN, Raymond LI, Christopher PAL, Jonathan PILAULT
-
Patent number: 11755909Abstract: There is provided a method and a system for training an extractive machine learning algorithm (MLA) to generate extractive summaries of text documents. Reference documents and associated extractive summaries are received. The extractive MLA is then trained to generate an extractive summary, where the training includes, for a given reference document, encoding, using a sentence encoder, a plurality of reference sentences to obtain an associated plurality of sentence representations, encoding, using a document encoder, the associated plurality of sentence representations to obtain a document representation, extracting, using a decoder and based on the associated plurality of sentence representations and the document representation, a first reference sentence of the plurality of reference sentences to obtain a first extracted sentence. A given parameter is updated based on the first extracted sentence and the given reference document summary.Type: GrantFiled: June 7, 2022Date of Patent: September 12, 2023Assignee: ServiceNow Canada Inc.Inventors: Sandeep Subramanian, Raymond Li, Christopher Pal, Jonathan Pilault
-
Publication number: 20220366251Abstract: There is provided a method and a system for training an extractive machine learning algorithm (MLA) to generate extractive summaries of text documents. Reference documents and associated extractive summaries are received. The extractive MLA is then trained to generate an extractive summary, where the training includes, for a given reference document, encoding, using a sentence encoder, a plurality of reference sentences to obtain an associated plurality of sentence representations, encoding, using a document encoder, the associated plurality of sentence representations to obtain a document representation, extracting, using a decoder and based on the associated plurality of sentence representations and the document representation, a first reference sentence of the plurality of reference sentences to obtain a first extracted sentence. A given parameter is updated based on the first extracted sentence and the given reference document summary.Type: ApplicationFiled: June 7, 2022Publication date: November 17, 2022Applicant: ServiceNow Canada Inc.Inventors: Sandeep SUBRAMANIAN, Raymond LI, Christopher PAL, Jonathan PILAULT
-
Patent number: 11397892Abstract: A method and a system for generating an abstractive summary of a document using an abstractive machine learning algorithm (MLA) and a method and a system for training the abstractive MLA. A document including a plurality of text sequences is received. An extractive summary of the document is generated, the extractive summary including a set of summary text sequences which is a subset of the plurality of text sequences. The abstractive MLA generates, based on the set of summary text sequences and at least a portion of the plurality of text sequences, an abstractive summary of the document including a set of abstractive text sequences, at least one abstractive text sequence not being included in the plurality of text sequences. In some aspects, the extractive summary is generated by an extractive MLA having been trained to generate extractive summaries.Type: GrantFiled: May 22, 2020Date of Patent: July 26, 2022Assignee: SERVICENOW CANADA INC.Inventors: Sandeep Subramanian, Raymond Li, Jonathan Pilault, Christophe Pal
-
Publication number: 20210365773Abstract: A method and a system for generating an abstractive summary of a document using an abstractive machine learning algorithm (MLA) and a method and a system for training the abstractive MLA. A document including a plurality of text sequences is received. An extractive summary of the document is generated, the extractive summary including a set of summary text sequences which is a subset of the plurality of text sequences. The abstractive MLA generates, based on the set of summary text sequences and at least a portion of the plurality of text sequences, an abstractive summary of the document including a set of abstractive text sequences, at least one abstractive text sequence not being included in the plurality of text sequences. In some aspects, the extractive summary is generated by an extractive MLA having been trained to generate extractive summaries.Type: ApplicationFiled: May 22, 2020Publication date: November 25, 2021Applicant: Element AI Inc.Inventors: Sandeep SUBRAMANIAN, Raymond LI, Jonathan PILAULT, Christophe PAL
-
Publication number: 20210134173Abstract: A method, system, and storage device storing a computer program, for generating questions based on provided content, such as, for example, a document having words. The method comprises automatically estimating the probability of interesting phrases in the provided content, and generating a question in natural language based on the estimating. In one example embodiment herein, the estimating includes predicting the interesting phrases as answers, and the estimating is performed by a neural model. The method further comprises conditioning a question generation model based on the interesting phrases predicted in the predicting, the question generation model generating the question. The method also can include training the neural model. In one example, the method further comprises identifying start and end locations of the phrases in the provided content, and the identifying includes performing a dot product attention mechanism parameterizing a probability distribution.Type: ApplicationFiled: January 7, 2021Publication date: May 6, 2021Applicant: Microsoft Technology Licensing, LLCInventors: Xingdi YUAN, Tong WANG, Adam Peter TRISCHLER, Sandeep SUBRAMANIAN
-
Patent number: 10902738Abstract: A method, system, and storage device storing a computer program, for generating questions based on provided content, such as, for example, a document having words. The method comprises automatically estimating the probability of interesting phrases in the provided content, and generating a question in natural language based on the estimating. In one example embodiment herein, the estimating includes predicting the interesting phrases as answers, and the estimating is performed by a neural model. The method further comprises conditioning a question generation model based on the interesting phrases predicted in the predicting, the question generation model generating the question. The method also can include training the neural model. In one example, the method further comprises identifying start and end locations of the phrases in the provided content, and the identifying includes performing a dot product attention mechanism parameterizing a probability distribution.Type: GrantFiled: August 3, 2017Date of Patent: January 26, 2021Assignee: Microsoft Technology Licensing, LLCInventors: Xingdi Yuan, Tong Wang, Adam Peter Trischler, Sandeep Subramanian
-
Publication number: 20190043379Abstract: A method, system, and storage device storing a computer program, for generating questions based on provided content, such as, for example, a document having words. The method comprises automatically estimating the probability of interesting phrases in the provided content, and generating a question in natural language based on the estimating. In one example embodiment herein, the estimating includes predicting the interesting phrases as answers, and the estimating is performed by a neural model. The method further comprises conditioning a question generation model based on the interesting phrases predicted in the predicting, the question generation model generating the question. The method also can include training the neural model. In one example, the method further comprises identifying start and end locations of the phrases in the provided content, and the identifying includes performing a dot product attention mechanism parameterizing a probability distribution.Type: ApplicationFiled: August 3, 2017Publication date: February 7, 2019Applicant: Microsoft Technology Licensing, LLCInventors: Xingdi YUAN, Tong WANG, Adam Peter TRISCHLER, Sandeep SUBRAMANIAN