全文下载：2013030305
张炯1，2
(1.山东商业职业技术学院 信息技术学院,山东 济南 250103；
2.国家农产品现代物流工程技术研究中心,山东 济南 250103）
摘要： 学术web平台蕴含着巨大的信息，通过统计分析提取这些信息,以便从海量非结构化数据世界中挖掘出新颖的、潜在的有用模式,正在成为数据分析处理的研究方向。针对已分大类的学术网页描述提取词汇链，创新提出最大相关方法对其进行扩充形成新的特征词是发现专业研究新趋势的有效手段。通过评价实验发现对词汇链拓展的最大相关方法比TF-IDF方法、最大熵方法、词汇链方式提取的特征词或短语更能有效地反映该研究方向相关趋势。
关键词： 学术网络平台； 共现特征词； 最大关联度； 特征提取
中图分类号： TP 391.1文献标志码： A
A Method for Discovering the Academic Network Platform of Technology Trends
ZHANG Jiong 1,2
(1．School of Information, Shandong Institute of Commerce and Technology, Ji’nan 250103，China; 2．National Engineering Research Center for Agricultural Products Logistics, Ji’nan 250103，China)
Abstract: Academic web platform contains great information, extracting these information through statistical analysis to dig out the novel, potentially useful pattern from massive unstructured data world, is becoming the research direction of data analysis and processing. It is an effective way to find new trend of professional research by extracting lexical chain in large class academic webpage description and innovatively proposing maximal correlation method to expand to form the new characteristic words. It is found through evaluation experiments that the characteristic words or phrases extracted from the maximal correlation method based on lexical chain expansion are more effective on research trend prediction compared with TF-IDF method and maximum entropy method.
Key words:academic network platform; concurrent words；maximum association；key terms extraction
收稿日期:2012-12-14
基金项目: 国家高技术研究发展计划（863计划）项目（2011AA100702）.
作者简介: 张炯(1972—)，男，讲师.
