时间序列下关键词多词共现分析及研究热点预测——以我国图情领域核心期刊为例  

Multiple-keyword Co-occurrence Analysis and Research Hotspots Prediction Based on Time Series: Case Study of Core Journals in Library and Information Science in China

作　　者：肖明 徐烨 Xiao Ming;Xu Ye(School of Government,Beijing Normal University,Beijing 100875)

机构地区：[1]北京师范大学政府管理学院

出　　处：《情报探索》2019年第11期115-120,共6页Information Research

基　　金：国家社科基金项目“基于语义识别的引文分析理论、方法与应用研究”(项目编号：16BTQ073)研究之一

摘　　要：[目的/意义]提出挖掘关键词多词共现的方法,从而揭示领域的研究主题并预测未来的研究热点。[方法/过程]以我国图情领域为例,通过Apriori算法,对近12年15种核心期刊关键词矩阵进行关联规则挖掘,并结合时间序列预测法对挖掘出的关键词集进行短期预测。[结果/结论]该方法挖掘出442组关键词集,得出近12年来我国图情领域的研究热点,并选择了4组典型的关键词集进行短期预测,分析这些领域的未来趋势。[Purpose/significance]The paper proposes a method to mine the co-occurrence of keywords,so as to reveal the research topics in the field and predict future research hotspots.[Method/process]The paper takes the field of library and information science in China as an example,uses Apriori algorithm to mine association rules for 15 kinds of core journal keyword matrices in recent 12 years,then makes short-term prediction for the keyword sets mined by using time series prediction method.[Result/conclusion]The method mines a total of 442 keyword sets,and obtains the research hotspots in the field of library and information science in recent 12 years,and selects 4 typical keyword sets to predict their support,and then analyzes future trends in these areas.

关 键 词：多词共现 APRIORI算法 时间序列预测法 关键词集 研究热点 

分 类 号：G250[文化科学—图书馆学]