摘要
[目的 /意义]将闭频繁项集挖掘技术应用于专利文本,进而分析技术的演化发展趋势,从微观层次描绘既定领域中主要的技术发展脉络。[方法 /过程]在进行闭频繁项集挖掘过程中,以专利文本的术语为事务、术语中的单词作为项,继而使用闭频繁项集之间的关联规则建立起术语层次网络,最后以术语层次网络为依托创建技术路线图。[结果 /结论]实证结果表明,该方法应用于硬盘驱动器磁头领域,能够刻画出不同术语的融合、派生过程,从而描绘出更加符合技术演化真实情况的技术路线图。
[ Purpose/significance ] This paper applies closed frequent itemsets on patent text for highlight of main technology paths on certain field. [ Method/process ] During mining frequent closed itemsets, we take terminologies as transaction, take words in terminologies as items, then extract association rules among these closed frequent itemsets to generate technology roadmap. [ Result/conclusion ] We apply this method on technology of hard disk drive head, and the empirical result shows that it can describe the merging and splitting process among different terminologies, thus enable to generate technology roadmap which is more approximate to reality.
出处
《图书情报工作》
CSSCI
北大核心
2016年第6期70-76,共7页
Library and Information Service
基金
国家科技支撑计划课题"专利信息资源整合与加工关键技术与规范研究"(课题编号:2013BAH21B01)研究成果之一
关键词
专利分析
技术演化
闭频繁项集
关联规则挖掘
硬盘磁头
patent analysis
technological evolution
frequent closed itemset
association rule mining
thin film head
