Patents by Inventor Arpita Roy

Arpita Roy 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).

  • Patent number: 11132507
    Abstract: A first vector representation of a first word within a first narrative text and a machine-generated label corresponding to the first word are constructed. Using the first vector representation, an annotator model is trained. The annotator model is configured to produce a set of probabilities, each probability in the set of probabilities representing a probable output annotation corresponding to a word within a narrative text. The training includes minimizing a difference between a first human-generated label corresponding to the first word and a first probable output annotation corresponding to the first word. Using the trained annotator model and a second narrative text, second training data is generated. The trained annotator model is configured to produce an output annotation corresponding to a word within a narrative text. The second training data is usable to train a relation extraction model.
    Type: Grant
    Filed: April 2, 2019
    Date of Patent: September 28, 2021
    Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATION
    Inventors: Youngja Park, Taesung Lee, Arpita Roy
  • Publication number: 20200320171
    Abstract: A first vector representation of a first word within a first narrative text and a machine-generated label corresponding to the first word are constructed. Using the first vector representation, an annotator model is trained. The annotator model is configured to produce a set of probabilities, each probability in the set of probabilities representing a probable output annotation corresponding to a word within a narrative text. The training includes minimizing a difference between a first human-generated label corresponding to the first word and a first probable output annotation corresponding to the first word. Using the trained annotator model and a second narrative text, second training data is generated. The trained annotator model is configured to produce an output annotation corresponding to a word within a narrative text. The second training data is usable to train a relation extraction model.
    Type: Application
    Filed: April 2, 2019
    Publication date: October 8, 2020
    Applicant: International Business Machines Corporation
    Inventors: Youngja Park, Taesung Lee, Arpita Roy
  • Publication number: 20170045690
    Abstract: The disclosure features optical scramblers that can be used with spectrographs and other optical measurement systems, including Doppler-based radial velocity measurement systems and other precision spectroscopy. In one aspect, an optical scrambler includes a first fiber configured to receive measured light from an optical collection system, a second fiber configured to deliver the measured light to a detection system, and an optical coupling element positioned between a light output surface of the first fiber and a light input surface of the second fiber and configured to deliver the measured light from the first fiber to the second fiber. The coupling element defines an output focal position for the measured light that is delivered to the second fiber and the output focal position is located within 50 microns of an output surface of the coupling element.
    Type: Application
    Filed: August 11, 2016
    Publication date: February 16, 2017
    Inventors: Suvrath Mahadevan, Arpita Roy, Sam Halverson