SAO 기반의 의미론적 특허 유사성을 활용한 특허맵 생성방법 = Generation of Patent Maps using SAO-based Semantic Patent Similarity

국문 초록 (Abstract)

특허는 기술의 진보를 반영하는 신뢰성 있는 최신 정보의 보고이며, 따라서 특허분석은 기술동향의 분석과 기술전략의 수립에 필수적인 요소로 인식되어 왔다. 특허분석을 위한 도구 중의 하나인 특허맵은 특허들간의 관계를 가시화함으로써, 특허공백, 특허클러스터, 특허경쟁현황 등 특허기반의 다양한 분석을 가능하도록 한다. 지금까지 특허맵은 키워드의 출현, 인용관계, 기술카테고리 등을 활용하여 작성되어 왔지만, 이들은 특허의 발명원리의 의미적 유사성을 반영하지 못한다. 따라서 본 연구에서는 Subject-Action-Object(SAO) 기반의 의미론적 특허 유사성 분석을 활용한 새로운 특허맵 작성방법을 제시하고자 한다. 특허 텍스트의 문장구조 분석을 통해 추출될 수 있는 SAO 구조는 기술을 구성하는 요소들 간의 연관관계를 명시적으로 표현함으로써 특허의 핵심개념 또는 발명자의 노하우를 담고 있는 핵심정보이다. 제시되는 방법은 1)특허정보의 수집, 2)자연어처리를 활용한 특허 텍스트로부터의 SAO 구조 추출, 3)각 특허에 포함된 SAO 구조들 간의 의미론적 문장 유사성 분석을 통해 특허 간의 비유사성 행렬을 생성, 4)다차원 척도법을 활용하여 비유사성 행렬을 2차원상에 매핑함으로써 특허맵을 생성한다. 본 연구에서 제시되는 방법은 실리콘 기반의 태양전지에 관련된 특허들을 활용하여 소개된다.

다국어 초록 (Multilingual Abstract)

Patents are an up-to-date technological knowledge source that reflect technological advance, and therefore patent analysis has been a vital component for technological trend analysis and technology strategy formulation. As one tool for patent analysis, patent maps visualize the overall relationship between patents and allow various patent-based analyses such as patent vacuum identification, patent infringement analysis and patent cluster analysis. Although there has been the patent maps based on keyword occurrence, citation or technology category, they do not show the internal similarity between patents. Therefore, this paper proposes a new patent map using Subject-Action-Object(SAO)-based semantic patent similarity measurement. In a patent, these ordered syntactic structures explicitly represent the relationship between technological factors and are even considered as the invention key concepts or the inventor’s know-how. The patent map is developed through 1)collecting patents, 2)extracting SAO structures from patent text using natural language processing, 3)generating a patent dissimilarity matrix by measuring semantic similarities between SAO structures of patents, and 4)visualizing the matrix using multidimensional scaling. The proposed method is illustrated using patents related to silicon-based thin film solar cells.

참고문헌 (Reference)

1 Miller, G.A., "WordNet: a lexical database for English" 38 (38): 41-, 1995

2 Simpson, T, "WordNet-based semantic similarity measurement"

3 Prinston, "WordNet, a lexical database for English"

4 Small, H., "Visualizing science by citation mapping" 50 (50): 799-813, 1999

5 Wanner, L., "Towards contentoriented patent document processing" 30 (30): 21-33, 2008

6 Kostoff, R.N., "The use and misuse of citation analysis in research evaluation" 43 (43): 27-43, 1998

7 Stanford, "The Stanford Parser: A statistical parser"

8 Resnik, P., "Semantic similarity in a taxonomy: An information-based measure and its application to problems of ambiguity in natural language" 11 (11): 130-, 1999

9 Kopcsa, A., "Science and technology mapping: A new iteration model for representing multidimensional relationships" 49 (49): 7-17, 1998

10 Moehrle, M.G., "Patent-based inventor profiles as a basis for human resource decisions in research and development" 35 (35): 513-524, 2005

11 윤병운, "On the development of a technology intelligence tool for identifying technology opportunity" PERGAMON-ELSEVIER SCIENCE LTD 35 (35): 124-135, 200807

12 Cyram, "NetMiner version 3.0 for network analysis"

13 Cascini, G., "Natural language processing of patents and technical documentation" 3163 : 89-92, 2004

14 Borg, I., "Modern multidimen-sional scaling: Theory and applications" Springer Verlag 2005

15 Cascini, G., "Measuring patent similarity by comparing inventions functional trees IFIP International Federation for Information Processing" Springer 31-42, 2008

16 Lin, Dekang, "MINIPAR"

17 Fujii, A., "Introduction to the special issue on patent processing" 43 (43): 1149-1153, 2007

18 Schiffman, S.S., "Introduction to multidimen-sional scaling: Theory, methods, and applica-tions" Academic Press New York 1981

19 Bergmann, I., "Evaluating the risk of patent infringement by means of semantic patent analysis: the case of DNA chips" 38 (38): 550-562, 2008

20 Moehrle, M.G., "Developing acquisition strategies based on patent maps Proceedings of the 13th International Conference on Management of Technology" 1-9, 2004

21 Huang, M.H., "Constructing a patent citation map using bibliographic coupling: A study of Taiwan's high-tech companies" 58 (58): 489-506, 2003

22 Knowlegist, "Comprehensive Knowledge Mining Tool"

23 Peters, H.P.F., "Co-word-based science maps of chemical engineering. Part II: Representations by combined clustering and multidimensional scaling" 22 (22): 47-71, 1993

24 Peters, H.P.F., "Co-word-based science maps of chemical engineering. Part I: Representations by direct multidimensional scaling" 22 (22): 23-45, 1993

25 Lee, S., "An approach to discovering new technology opportunities:Keyword-based patent map approach" 29 (29): 481-497, 2009

26 Salton, G., "A vector space model for automatic indexing" 18 (18): 613-620, 1975

27 Yoon, B., "A text-mining-based patent network: Analytical tool for high-technology trend" 15 (15): 37-50, 2004