论文部分内容阅读
【目的】弥补传统方法在潜在合作关系挖掘中的缺陷和不足,提高潜在合作关系的挖掘效果。【方法】在分析简单计算法、最小值计算法与传统TFIDF算法缺陷和不足的基础上,提出改进TFIDF算法,并将其引入到潜在合作关系挖掘中。【结果】利用《北大中文期刊核心目录(2012年版)》中19种图书情报类期刊近5年情报学研究方法应用领域的论文作为样本数据,发现简单计算法与最小值计算法受到作者发文量影响较大,传统TFIDF算法的挖掘结果很难实现从潜在合作关系转化为现实合作关系,而改进TFIDF算法对此的满足度则表现得非常突出。【局限】改进TFIDF算法未考虑论文中作者之间的排名顺序对潜在合作关系的影响。【结论】通过将4种挖掘结果进行对比和评价,证明改进TFIDF算法较其他传统方法更科学、更具有优越性和实用价值。
【Objective】 To make up for the shortcomings and shortcomings of traditional methods in mining potential cooperative relationship and to improve the mining effect of potential cooperation. 【Method】 Based on the analysis of the shortcomings and deficiencies of the simple calculation method, the minimum calculation method and the traditional TFIDF algorithm, an improved TFIDF algorithm is proposed and introduced into the potential cooperation mining. 【Results】 Using the essays in the field of information science research of 19 kinds of library and information science journals in the core catalog of Peking University Chinese periodicals (2012 edition) for the past five years as the sample data, it was found that the simple calculation method and the minimum value calculation method were influenced by the authors’ The impact of the traditional TFIDF algorithm is hard to realize from the potential cooperation to the actual cooperation, and the improvement of the TFIDF algorithm is very satisfied with the performance of this very prominent. [Limitations] The improved TFIDF algorithm does not consider the impact of rankings among authors in the paper on potential partnerships. 【Conclusion】 By comparing and evaluating the four kinds of excavation results, it is proved that the improved TFIDF algorithm is more scientific, superior and practical than other traditional methods.