青刺果德温特专利的专业性主题聚类  

Professional subject clustering of Derwent Innovation Index for Prinsepia utilis Royle

作　　者：陈一 Chen Yi(School of Management and Economics,Kunming University of Science and Technology,Kunming,Yunnan 650000,China)

机构地区：[1]昆明理工大学管理与经济学院,云南昆明650000

出　　处：《计算机时代》2022年第7期33-35,共3页Computer Era

基　　金：国家自然科学地区基金项目“青刺果多糖修复表皮通透屏障的分子机制研究”(81960744)。

摘　　要：专利文本会涉及不同学科领域的专业词汇。本文以青刺果德温特专利数据为例,研究专利的专业性主题聚类。运用Word2vec对青刺果德温特专利文本训练词向量,再用N-Gram改进TF-IDF优化加权,然后用LDA做主题聚类。结果显示,青刺果专利的研究热点主要集中在青刺果的功能、应用、提取技术和成分分析方面,种植和包装方面的相关专利需要加强。Patent texts involve professional vocabulary in different disciplines.In this paper,taking Derwent Innovation Index for Prinsepia utilis Royle as an example,the professional subject clustering of patents is studied.Firstly,Word2vec is used to train word vector of the patent text,then N-Gram is used to improve TF-IDF optimization weighting,and finally LDA is used for topic clustering.The results show that the research hotspots of the patents mainly focus on the function,application,extraction technology and component analysis of the Prinsepia utilis Royle,and the related patents in planting and packaging need to be strengthened.

关 键 词：专业性主题聚类 青刺果德温特专利 Word2vec N-Gram-TF-IDF LDA 

分 类 号：TP39[自动化与计算机技术—计算机应用技术] G35[自动化与计算机技术—计算机科学与技术]