학술논문

Promising Technology Selection based on the Use of Integrated Delphi, AHP and Patent Analysis Model : Application in Automobile Parts Industry
델파이, AHP 및 특허 분석의 통합 활용 모델에 기반한 유망 기술선정 : 자동차 부품 산업에의 적용
Document Type
Article
Text
Source
기업경영연구, 12/31/2012, Vol. 19, Issue 6, p. 283-303
Subject
델파이
계층분석방법
K-means 알고리즘
특허맵
자동차부품산업
Delphi Method
Analytic Hierarchy Process
K-means Algorithm
Patent Map
Automobile Parts Industry
Language
English
ISSN
1229-957X
Abstract
Nowadays, technology innovation is the important resource of national and industrial competitiveness. Forecasting correct technology R&D direction is very important for establishing correct technology strategy. Technology forecasting is an effective method for setting technology strategies. Technology forecasting method can help the decision makers to select promising technologies. In this study, we proposed a integrated Delphi, analytic hierarchy process (AHP) and patent analysis model to forecast promising technology and select optimal promising technology. In the proposed model, Delphi method is used to select the technology alternatives which have high ratings at importance and urgency. AHP is used to prioritize the selected technology alternatives considering importance, specialty and urgency. In the meantime, patents of the technology alternatives are collected. Then, keywords of the collected patents are clustered by K-means algorithm, registration date of the collected patents are compared to create the patent maps which can display the technology’s development trend. Finally, the proposed model compares the technology alternative’s priority and the patent map, considers the association between the alternatives and the keywords, and select alternatives which have high priority and meet the future development trend as the promising technologies. For illustration, we applied the proposed model to the automobile parts industry’s electric apparatus technology to forecast and select promising electric apparatus technologies. The application result showed that the proposed model can effectively forecast promising technologies and select optimal promising technologies.