面向学术社区的专家推荐模型

发布时间：2018-03-31 15:51

本文选题：学术专家推荐　切入点：H参数　出处：《智能系统学报》2012年04期

【摘要】：在学术社区提供的服务中,对于研究者特别是青年研究者来说,专家推荐是一个必不可少的部分.目前提供学术信息服务的所有中文搜索引擎中,都没有提供用户感兴趣的专家推荐服务.因此,提出了一个面向学术社区的专家推荐模型.使用改进的H参数对学者n年时间内发表的论文成果进行量化,获取专家列表;使用概率主题模型从作者发表的论文中提取主题向量作为学者的研究方向;根据矩阵奇异值分解对构建的词项-文档矩阵进行降维,进而生成词项-词项关系矩阵,实现对搜索关键词的查询扩展,并计算查询扩展向量与作者主题向量之间的相关度,根据相关度大小进行排序推荐.在SCHOLAT(学者网)数据集上验证模型的有效性,实验结果表明提出的模型达到了预期的效果.
[Abstract]:Expert recommendation is an essential part of the services provided by the academic community, especially young researchers. Therefore, an expert recommendation model for the academic community is proposed. The improved H parameter is used to quantify the results of the papers published within a period of n years, and the list of experts is obtained. Using the probabilistic topic model to extract the topic vector from the paper published by the author as the research direction, according to the singular value decomposition of the matrix to reduce the dimension of the word-document matrix, and then to generate the word-to-word relationship matrix. The query expansion of search keywords is realized, and the correlation between the query expansion vector and the author topic vector is calculated, and the ranking recommendation is made according to the correlation degree. The validity of the model is verified on the Scholarnet data set. The experimental results show that the proposed model achieves the desired results.
【作者单位】：肇庆学院计算机学院;华南师范大学计算机学院;广东技术师范学院计算机学院;
【基金】：国家自然科学基金资助项目(60970044) 广东省科技计划资助项目(2010B010600031) 广州市科技计划资助项目(2010J-D00511)
【分类号】：TP393.09

【参考文献】