基于引用情感交互的学术检索结果排序方法研究

Research on the Ranking Method of Academic Retrieval Results Based on Citation Sentiment Interaction

[目的/意义]传统学术检索系统常从被引次数、发表时间、标题的相关性等单一角度对检索结果进行排序,而忽略了文献引证关系以及引用内容中蕴含的情感观点造成的文献内在价值波动,导致文献排名区分度不够。[方法/过程]针对这些问题,国内外很多学者提出将PageRank算法应用到文献检索中,从引证关系角度对排序结果进行改进。文章希望在考虑引证关系的同时将引用情感也融入其中,从引用内容着手根据文献的内在价值进行排序。[结果/结论]实验证明,基于引用情感交互的检索排序方法比传统基于被引次数以及PageRank的排序方法效果更好,能综合考虑多种因素对文献内在价值的影响,从而使得文献的价值判断也更为客观和准确。

[Purpose/significance]The traditional academic retrieval system often sorts the search results from a single angle such as citation frequency,publication time,or title relevance,while ignoring the literature citation relationship and the intrinsic value fluctuation of the literature brought by the emotional opinions contained in the citation content while resulting in insufficient discrimination in the literature ranking.[Method/process]In response to these problems,many domestic and foreign scholars have proposed applying the PageRank algorithm to literature retrieval,and improving the ranking results from the perspective of citation.This paper hopes to incorporate citation emotions while considering the citation relationship,so can rank the intrinsic value of the literature from the perspective of citation content.[Result/conclusion]The experiment shows that the retrieval ordering method based on citation emotional interaction is better than the traditional ranking method based on the number of citations or PageRank algorithm.It can comprehensively consider the influence of various factors on the intrinsic value of the literature,which makes the value judgment of the literature more objective and accurate.