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

오용환 (경북대학교, 경북대학교 대학원)

지도교수
박성배
발행연도
2015
저작권
경북대학교 논문은 저작권에 의해 보호받습니다.

이용수0

표지
AI에게 요청하기
추천
검색

이 논문의 연구 히스토리 (2)

초록· 키워드

오류제보하기
A content-based filtering recommends items based on the comparison between the content of the items and a user preference. The user preference is represented with the content which have been seen by the user. Since the performance of recommendation is influenced by the quality of the contents, contents must be represented well.
App recommendation is a task to recommend the most suitable app to a user. It can be performed with the content-based filtering by regarding the content as an app description and the user preference as descriptions of apps used by the user. However, the app description includes not only a functionality content but also a non-functionality content that is not related with the app. When we use the app description as the content, the item and user preference are biased from the non-functionality contents of the app description. As a result, it requires to eliminate the non-functionality contents.
This paper proposes a method to improve the performance of app recommendation. In general, the app description consists of several paragraphs and each paragraph represents the same theme. By removing paragraphs which contain non-functionality content, we can obtain the only functionality content. It can be formulated as a binary classification, that is, given a paragraph in the app description, we classify it as whether non-functionality content or not. Then, we use the only functionality content of app description to represent the app and the user preference.
In experiments, we first check how the proposed method eliminates the non-functionality contents well. It achieves 93.25% accuracy. This result implies that the proposed method is appropriate to remove the non-functionality contents. Second, the performance of app recommendation used only the functionality contents of app description achieves higher performance than the one used both functionality and non-functionality contents. The result proved that when we recommend the app based on the content-based filtering, the non-functionality contents of app description should be eliminated.

목차

1. 서 론 1
2. 관련 연구 6
3. 앱 설명의 노이즈 제거를 통한 모바일 앱 추천 8
3.1 앱 설명의 노이즈 제거 9
3.2 모바일 앱 추천 18
4. 실 험 22
4.1 앱 설명의 노이즈 제거 성능 22
4.1.1 실험 데이터 22
4.1.2 실험 및 평가 25
4.2 모바일 앱 추천 성능 27
4.2.1 실험 데이터 27
4.2.2 실험 및 평가 30
5. 결 론 35
참고문헌 37
영문초록 40

최근 본 자료

전체보기

댓글(0)

0