Abstract: Based on the current situation of co-word network analysis which could be applied to public policy, collected innovation policies from National Innovation Demonstration Zone(NIDZ) as the research objection, the author discusses the influences and the functions of policies attributes and key words weights in the co-word analysis. Then, the author uses grounded theory, word frequency and policy measurement, to model the weighted co-word analysis based on the weightes of policies and key words. After building up the model of weighted co-word analysis, the author calculates the distances between keywords by weighted Ochiia Inex. Through node analysis and group analysis to validate the weighted co-word model, and to demonstrate the current situation and deficiencies from comprehensive viewing angle, to find technological finance, industry development and talent training still have room to improvement, and to support policy-making in the future.