PATSTAT(WorldwidePatentStatisticalDatabase,全球专利统计数据库)是一款由EPO欧洲专利局开发的、面向统计决策的专利数据仓库。该数据库在专利数据集成方面取得一系列的研究进展,其在语义映射、语义匹配方面的经验对于解决现阶段专利数据语义异构性问题具有较强的参考意义。通过逆向归纳的方法,文章阐述PATSTAT数据库在专利家族、优先权、摘要、标题、发明人、地址信息以及专利权人信息方面的集成策略,并对PATSTAT数据库在专利数据集成经验进行总结。
PATSTAT (Worldwide Patent Statistical Database) is the world's leading patent warehouse product developed by the EPO to serve statistical decision supporting. The product has made a series of research progress at patent data integration, and has instruction and reference value to solve the issues on patent semantic- level heterogeneity. The paper adopts a reverse inductive method, summarizing the progresses of PATSTAT database. Finally, this article summarizes the patent data integration experiences of PATSTAT database, which can use for patent data processing and analysis in future.