메뉴 건너뛰기
.. 내서재 .. 알림
소속 기관/학교 인증
인증하면 논문, 학술자료 등을  무료로 열람할 수 있어요.
한국대학교, 누리자동차, 시립도서관 등 나의 기관을 확인해보세요
(국내 대학 90% 이상 구독 중)
로그인 회원가입 고객센터 ENG
주제분류

추천
검색

논문 기본 정보

자료유형
학위논문
저자정보

신용민 (경북대학교, 경북대학교 대학원)

지도교수
박세영
발행연도
2020
저작권
경북대학교 논문은 저작권에 의해 보호받습니다.

이용수1

표지
AI에게 요청하기
추천
검색

이 논문의 연구 히스토리 (3)

초록· 키워드

오류제보하기
Recently, with the development of the Internet, people are exposed to a large amount of information. People have to extract information from many documents to get the information they want. That is why document summarization is needed. As the demand for high quality summary increases, many studies use abstractive summarization. The method using deep neural network model is best performance in single-document summarization. However, in multi-document summarization, it is difficult to train a deep neural network model due to the lack of data. So, many researches try to solve this problem by transfer learning from trained model with huge data like single document summarization data. However, these have some problem like input format between single document and multi- document and number of parameters to learned. In this paper, we propose two transfer learning methods from single document summarization model to multi document summarization model. One method considers the input format of a single document and multi document. Another method takes into account two things: 1) the number of trainable parameter, 2) lack of multi document summarization data. Each method is called Transfer Learning using Multi Document Encoder, and Transfer Learning using Adapter, respectively. To show the performance of our proposed method, we conducted experiments on DUC(Document Understand Conference) and multi-news datasets. As of Korean, because there was no public data, experiments were conducted by creating a new dataset. Both our methods demonstrate competitive results, and outperform previous researches. With a small amount of data are confirm that, Transfer Learning using Adapter shows excellent performance.

목차

Ⅰ. 서 론 1
Ⅱ. 관련 연구 5
Ⅲ. 문서 요약 모델 9
3.1. 단일 문서 요약 모델 9
3.2. 다중 문서 인코더를 이용한 전이학습 11
3.3. 어댑터를 이용한 전이학습 16
Ⅳ. 실 험 20
4.1. 데이터 구성 20
4.2. 실험 구성 23
4.3. 실험 결과 및 분석 25
Ⅴ. 결 론 36
참고문헌 37
영문초록 44

최근 본 자료

전체보기

댓글(0)

0