中文专利信息资源深加工方案设计与实证研究

Solution Design for Deep Processing of Chinese Patent Resources and Its Empirical Study

基于专利分析的视角，调研了国内外专利信息资源的现状和特点，明确了用于专利分析的资源及其加工深度的不足，设计了预处理和深加工的方案，重点为专利清洗和专利中四类信息的内容标引：发明类型、技术主题、发明改进和应用领域。并以新能源汽车领域为例，对中文专利进行了深加工实证研究。结果显示，利用清洗和深度标引后的专利资源能有效提高分析的准确度和专利资源揭示深度。

From patent analysis perspective, this paper surveys the present situation of patent resources and characteristics, and summarizes the shortcomings of patent resources used in analysis. Then it proposes a solution to process patent resource deeply. The emphases are cleaning and indexing of four type information：invention type, technology topic, improvement and application area. At last, it presents an empirical study, which proves that the cleaned and extracted patent resources are more accurate and revealed for analysis.