비정형데이터의 정형화 분석에 기반한 Fuzzy 추론 시스템

Abstract

Recent social and economic systems applying artificial intelligence have been actively studied, and the development of social and economic systems is directly related to the performance of data analysis. Accurate and reliable data analysis are used to interpret meaningful information in various environments, Based on this prediction, it can play a decisive role by providing information to important policy decisions such as disaster situations and facility maintenance. In general, data are classified into structured data and unstructured data. In the case of structured data has been developed using various kinds of analysis methods, but unstructured data has not. Because it requires much endeavor to analyze significance. In particular, unstructured data must be formatted to analyze it, and there is a problem in that it is necessary to precede the artificial classification without automatization in the structured data part. In this paper, propose a fuzzy inference system that it classify documents using LSA (Latent Semantic Analysis) in the analysis of such unstructured data, formulates unstructured data based on the classified data, and deduces the results using final formulated data. In addition, this system is standardized by applying context awareness, so that it can be used in various environments. The proposed system tests crime profiling as an example to confirm the effectiveness of unstructured data analysis and Finally, it is reflected in ITS (Intelligent Transport System) to deduce accurate and reliable traffic information and provide it to others.