论文部分内容阅读
对药物专利中族性结构进行标引是将专利文本信息转换成结构信息用于族性结构的匹配检索。本文根据国家知识产权局知识产权出版社数据研发中心提出的新的图形化标引语言GFL对近千余篇药物专利进行了标引。结合GFL的特点,统计分析药物专利中出现频率很高的结构片断,将这些结构片断的标引结果利用面向对象编程技术设计成一系列的族性模板,便于标引人员直接调用,有利于实现标引工作的半自动化。
The indexing of clan structures in drug patents is to convert the patent text information into structural information for matching retrieval of clan structures. In this paper, based on the new Graphical Citation Language (GFL) proposed by the Data Research and Development Center of Intellectual Property Press of State Intellectual Property Office, nearly one thousand patent titles are cited. Combined with the characteristics of GFL, statistical analysis of drug patents in the high frequency of structural fragments, the structure of these fragments of the indexing results using object-oriented programming techniques designed into a series of ethnic templates, to facilitate direct call indexing staff is conducive to the realization of standard Guide the work of semi-automation.