유사도를 활용한 신뢰성 있는 저자 정보 확장 방법 = A Method for Expanding Reliable Author Information Using Similarity

다국어 초록 (Multilingual Abstract)

The field of linked data is being actively pursued by the research community to explore new methods for connecting and expanding data. In order to continue connecting data, it is necessary that similar objects are identified. But to the best of our knowledge, so far no reliable object recognition method has been proposed. In this paper we proposes a new similarity calculation method. Our proposed method is based on public confidence level of web documents. This method will help in identification of objects (authors) and thereby pave the way for integration of related URI’s that will be finally presented to users. Our dataset used for verification, includes details about 10 objects (authors) to calculate recall and precision. This study compared with previous studies can attain higher accuracy results.

목차 (Table of Contents)

• 1. 서 론 1
• 1.1 연구의 배경 1
• 1.2 연구의 목적 3
• 1.3 연구 방법 및 구성 4
• 2. 관련 연구 5
• 2.1 링크드 데이터 5
• 2.2 sameAs 8
• 2.3 유사도 척도 10
• 2.4 TF/IDF 13
• 3. 저자 정보 확장 방법 14
• 3.1 온톨로지 구성 15
• 3.2 데이터 수집 단계 16
• 3.3 계산 및 측정 단계 19
• 3.3.1 가중치 적용 및 저자 식별 20
• 3.4 온톨로지 연결 및 통합 24
• 4. 실험 26
• 4.1 실험 방법 26
• 4.2 실험 결과 및 평가 29
• 5. 결론 및 향후 연구 31
• 5.1 결론 31
• 5.2 향후 연구 32
• 참 고 문 헌 33
• ABSTRACT 37