Title:
ELECTRONIC DEVICE AND METHOD FOR PROVIDING OUTPUT SENTENCE FROM INPUT SENTENCE
Document Type and Number:
WIPO Patent Application WO/2024/025020
Kind Code:
A1
Abstract:
A processor of an electronic device according to one embodiment may obtain a first data set representing words in a first sentence classified into a first category. The processor may input the first data set and parameters representing a second category into a first neural network so as to identify a second data set representing words related to the second category. The processor may obtain a second sentence on the basis of the second category by using the identified second data set. The first neural network may be trained, on the basis of a second neural network for obtaining a relationship between a third data set and the second category from the third data set outputted from the first neural network, and a third neural network for outputting a fourth data set representing a third sentence within the second category from specified words.
More Like This:
| JP2020047028 | INFORMATION PROCESSING SYSTEM, METHOD AND PROGRAM |
| JP2023120030 | INFORMATION PROCESSOR AND INFORMATION PROCESSING PROGRAM |
| WO/2021/230979 | INTELLIGENT LOCALIZATION OF RESOURCE DATA |
Inventors:
CHO HOJUN (KR)
CHOO JAEGUL (KR)
HWANG JEONGIN (KR)
NOH HYUNGJONG (KR)
CHOO JAEGUL (KR)
HWANG JEONGIN (KR)
NOH HYUNGJONG (KR)
Application Number:
PCT/KR2022/011255
Publication Date:
February 01, 2024
Filing Date:
July 29, 2022
Export Citation:
Assignee:
NCSOFT CORP (KR)
KOREA ADVANCED INST SCI & TECH (KR)
KOREA ADVANCED INST SCI & TECH (KR)
International Classes:
G06F40/58; G06F40/279; G06N3/04; G06N3/08
Other References:
DA-SOL PARK, JEONG-WON CHA: "Text Style Transfer of Non-parallel Data using Transformer and Discriminator", THE 32TH ANNUAL CONFERENCE ON HUMAN & COGNITIVE LANGUAGE TECHNOLOGY (HCLT 2020), SIGHCLT: SPECIAL INTEREST GROUP OF HUMAN AND COGNITIVE LANGUAGE TECHNOLOGY; OCTOBER 15, 2020 - FRIDAY, OCTOBER 16, 2020, KOREAN INSTITUTE OF INFORMATION SCIENTISTS AND E, 14 October 2020 (2020-10-14) - 16 October 2020 (2020-10-16), Korea, pages 64 - 68, XP009553185
YANG ZICHAO, HU ZHITING, DYER CHRIS, XING ERIC P, BERG-KIRKPATRICK TAYLOR: "Unsupervised Text Style Transfer using Language Models as Discriminators", ARXIV.ORG, CORNELL UNIVERSITY LIBRARY, ARXIV.ORG, ITHACA, 29 May 2018 (2018-05-29), Ithaca, XP093133064, Retrieved from the Internet [retrieved on 20240220], DOI: 10.48550/arxiv.1805.11749
DAI NING, LIANG JIANZE, QIU XIPENG, HUANG XUANJING: "Style Transformer: Unpaired Text Style Transfer without Disentangled Latent Representation", PROCEEDINGS OF THE 57TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, STROUDSBURG, PA, USA, 1 January 2019 (2019-01-01), Stroudsburg, PA, USA, pages 5997 - 6007, XP093133066, DOI: 10.18653/v1/P19-1601
JANG-SEONG BAE, CHANG-GI LEE, JEONG-IN HWANG, HYEONG-JONG NOH: "Unpaired Korean Text Style Transfer with Masked Language Model ", KOREAN SOCIETY OF INFORMATION SCIENTISTS AND ENGINEERS LANGUAGE ENGINEERING RESEARCH SOCIETY 2021 33RD HANGUL AND KOREAN INFORMATION PROCESSING CONFERENCE, KOREAN SOCIETY OF INFORMATION SCIENTISTS AND ENGINEERS LANGUAGE ENGINEERING RESEARCH SOCIETY, , 14 October 2021 (2021-10-14) - 14 October 2021 (2021-10-14), Korea, pages 391 - 394, XP009553661
LAI HUIYUAN, TORAL ANTONIO, NISSIM MALVINA: "Thank you BART! Rewarding Pre-Trained Models Improves Formality Style Transfer", PROCEEDINGS OF THE 59TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS AND THE 11TH INTERNATIONAL JOINT CONFERENCE ON NATURAL LANGUAGE PROCESSING (VOLUME 2: SHORT PAPERS), ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, STROUDSBURG, PA, U, 1 January 2021 (2021-01-01), Stroudsburg, PA, USA, pages 484 - 494, XP093133067, DOI: 10.18653/v1/2021.acl-short.62
YANG ZICHAO, HU ZHITING, DYER CHRIS, XING ERIC P, BERG-KIRKPATRICK TAYLOR: "Unsupervised Text Style Transfer using Language Models as Discriminators", ARXIV.ORG, CORNELL UNIVERSITY LIBRARY, ARXIV.ORG, ITHACA, 29 May 2018 (2018-05-29), Ithaca, XP093133064, Retrieved from the Internet
DAI NING, LIANG JIANZE, QIU XIPENG, HUANG XUANJING: "Style Transformer: Unpaired Text Style Transfer without Disentangled Latent Representation", PROCEEDINGS OF THE 57TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, STROUDSBURG, PA, USA, 1 January 2019 (2019-01-01), Stroudsburg, PA, USA, pages 5997 - 6007, XP093133066, DOI: 10.18653/v1/P19-1601
JANG-SEONG BAE, CHANG-GI LEE, JEONG-IN HWANG, HYEONG-JONG NOH: "Unpaired Korean Text Style Transfer with Masked Language Model ", KOREAN SOCIETY OF INFORMATION SCIENTISTS AND ENGINEERS LANGUAGE ENGINEERING RESEARCH SOCIETY 2021 33RD HANGUL AND KOREAN INFORMATION PROCESSING CONFERENCE, KOREAN SOCIETY OF INFORMATION SCIENTISTS AND ENGINEERS LANGUAGE ENGINEERING RESEARCH SOCIETY, , 14 October 2021 (2021-10-14) - 14 October 2021 (2021-10-14), Korea, pages 391 - 394, XP009553661
LAI HUIYUAN, TORAL ANTONIO, NISSIM MALVINA: "Thank you BART! Rewarding Pre-Trained Models Improves Formality Style Transfer", PROCEEDINGS OF THE 59TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS AND THE 11TH INTERNATIONAL JOINT CONFERENCE ON NATURAL LANGUAGE PROCESSING (VOLUME 2: SHORT PAPERS), ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, STROUDSBURG, PA, U, 1 January 2021 (2021-01-01), Stroudsburg, PA, USA, pages 484 - 494, XP093133067, DOI: 10.18653/v1/2021.acl-short.62
Attorney, Agent or Firm:
KWANG AND JANG INTELLECTUAL PROPERTY LAW FIRM (KR)
Download PDF: