Abstract:

Policy text computing is the product of the integration of information science, policy science and computational science. Through policy text computing, we analyze and study the contents of policy documents and policy reviews, making the policy making process more scientific and effective. First, we use keyword extraction to extract policy framework from policy documents. Secondly, cluster analysis of policy review data is conducted by thematic analysis. Thirdly, co-word analysis method is used to computing the intensity data of the collected policy documents and policy reviews. Finally, the policy contents are analyzed based on the computing data. By optimizing coding rules, the article improves the fineness of data extraction. The validity of the policy content analysis method based on the policy text computing is demonstrated by using the Internet renting bicycle as an example. This method provides a new way of thinking for the future policy analysis.

Key words: Policy text computing, Policy content analysis, Internet renting bicycles, Latent Dirichlet Allocation(LDA), Content analysis method