基于领域本体的专利技术-功效文本挖掘方法——以MOCVD技术为例  

A Method of Patent Technology-Function Mining Based on Domain Ontology: MOCVD as an Example

作　　者：陆佳伟 慎金花[1] 张更平[1] 杨锋[1] 

机构地区：[1]同济大学经济与管理学院,上海200092

出　　处：《价值工程》2018年第2期245-248,共4页Value Engineering

摘　　要：专利技术-功效图是专利地图的一种。通过制作专利技术-功效图,可以有效识别相关技术空白点和突破点,有助于加强专利战略部署。本文以MOCVD(金属有机化合物化学气相沉淀)技术为例,首先通过技术骨架的搭建确定技术-功效词,构建了一个技术-功效导向的基于叙词表的领域本体。进而借助文本挖掘技术通过专利摘要分析初步实现了对文本语义的理解,完成了技术-功效图的半自动制作。通过将研究结果与某公司人工制作的技术-功效图进行对比,验证了此方法的准确性和有效性,并对二者的差异进行了解释。Patent technology-function diagram is a kind of patent map. Through the production of patent technology-function diagram,we can effectively identify the relevant technical gaps and breakthrough points, help strengthen the patent strategy deployment. This paper takes MOCVD（Metal-organic Chemical Vapor Deposition） technology as an example, first to establish the technical words and efficacy words based technical backbone structures in order to build a technology and function-oriented domain ontology on the basis of thesaurus.And then by means of text mining patent abstract analysis, this method realizes text semantic understanding initially as well as figures out the technology-function diagram semi-automatically. By comparing with a company with its artificial production of technology-function diagram, the accuracy and effectiveness of this method is verified and the difference between the two is explained.

关 键 词：领域本体 专利摘要分析 技术-功效图 

分 类 号：G350.7[文化科学—情报学]