상세정보
학술대회
Tagged Video Retrieval System using Domain Ontology and Word Embedding
- 발행일
- 201710
- 출처
- International Conference on Information and Communication Technology Convergence (ICTC) 2017, pp.1101-1103
- 협약과제
- 17ZH9900, 오픈 시나리오 기반 프로그래머블 인터랙티브 미디어 창작 서비스 플랫폼 개발(이월액), 박종현
- 초록
- The number of users is interested in generating new contents by utilizing and editing existing videos. In order to utilize the existing videos, a video retrieval system is necessary to retrieve an appropriate source of videos. However, since the existing video retrieval system is based on keyword matching, there is a limitation to find videos they want. In this paper, we propose a query expansion method that can express users' search intention more accurately in order to solve these problems. The proposed method uses the domain ontology and word embedding results to add semantically related words. This allows the user to obtain satisfactory search results. We evaluate the proposed query expanding approach using the implemented retrieval system.
- KSP 제안 키워드
- Domain ontology, Keyword Matching, Number of users, Search intention, Search results, Semantically related words, Video retrieval, Word Embedding, query expansion, retrieval system