《计算机应用研究》|Application Research of Computers

一种改进的个性化查询引文推荐方法

Improved method for personalized query citation recommendation

免费全文下载 (已被下载 次)  
获取PDF全文
作者 李飞,张宏鸣,蔡晓妍,刘斌,郭蓝天
机构 1.西北农林科技大学 信息工程学院,陕西 杨陵 712100;2.西北工业大学 自动化学院,西安 710072
统计 摘要被查看 次,已被下载
文章编号 1001-3695(2019)08-010-2289-05
DOI 10.19734/j.issn.1001-3695.2018.02.0086
摘要 为充分利用文本内容的上下文信息,结合图模型及查询向量的构建方法,提出一种融合查询内容信息的个性化引文推荐方法。通过三种论文信息构建三层图模型,并在不同层上设置不同参数,调整节点向不同层次的跳转概率;利用word2vec技术构建的查询向量,可以有效利用文本上下文内容信息,使相似的文章在距离上更加接近,进而对候选文章进行评分预测与论文推荐。在association of computational linguistics anthology network数据集上进行计算分析,相同查询下与原有的方法相比在recall@<i>N</i>上平均提高约7%,在NDCG@<i>N</i>上平均提高约11%。实验结果表明该方法可以使引文推荐的质量得到有效的提升,能够获得较好的推荐效果。
关键词 多关系图; 词向量; 查询向量; 带重启的随机游走; 个性化推荐
基金项目 国家自然科学基金资助项目(41771315,41301283,41371274)
国家重点研发计划资助项目(2017YFC0403203)
欧盟地平线2020研究与创新计划资助项目(GA:635750)
陕西省自然科学基金面上项目(2017JM6059)
本文URL http://www.arocmag.com/article/01-2019-08-010.html
英文标题 Improved method for personalized query citation recommendation
作者英文名 Li Fei, Zhang Hongming, Cai Xiaoyan, Liu Bin, Guo Lantian
机构英文名 1.College of Information Engineering,Northwest A & F University,Yangling Shaanxi 712100,China;2.School of Automation,Northwestern Polytechnical University,Xi'an 710072,China
英文摘要 To make full use of the context information of the papers, combined with the graph model and the construction method of query vector, this paper proposed a fusion query information personalized citation recommendation method. It built a three layer graph model through three kinds of paper information, and set different parameters on different layers to adjust the jump probability of nodes to different levels; the query vector constructed using word2vec technology could effectively use the text context information, so that similar papers were closer to the distance, and then predicted and recommended the candidate papers. Computational analyzes performed on the association of computational linguistics anthology network dataset showed an average increase of about 7% over recall@<i>N</i> and an average increase of about 11% over NDCG@<i>N</i> for the same query compared to the original method. Experimental results show that the proposed method can effectively improve the quality of citation recommendation and get better recommendation results.
英文关键词 multi-relation; word vector; query vector; random walk with restarts; personalized recommendation
参考文献 查看稿件参考文献
 
收稿日期 2018/2/9
修回日期 2018/3/22
页码 2289-2293
中图分类号 TP301.6
文献标志码 A