言語統計分析による住宅建設五箇年計画及び答申の特性分析 : 政策の立案と評価における非定型・大量情報の活用可能性

概要

論文の詳細を見る
This paper illustrates the application of text mining approach to extracting potential information from HCFP (Housing Construction Five-year Program) text documents using lexical statistics. After obtaining the lexical statistic from the collected Reports (Toshin) and Program (Keikaku) of HCFP, we intensively investigated the relation between the actual documents' contextual meaning and the resulting information obtained by multivariate analysis and network analysis. The results clearly show that the lexical statistic of frequency and co-occurrence of key words have beneficial information on both capturing key issues and degree of centralities of key words in HCFP documents. These encouraging findings suggest the promising possibility of application of this methodology in urban planning and architecture domain.
社団法人日本建築学会の論文
2004-05-30