基于Vague相似性的文献推荐研究

Research on Literature Recommendation Based on Vague Similarity

为提高文献综合推荐效率,提出一种基于Vague相似性的文献推荐方法。该方法首先利用文献数据库检索用户感兴趣文献,作为待推荐文献集;然后,对文献中的关键词依据提出的方法重新排序、计算关键词的模糊隶属度和关键词转换为Vague值;最后,利用Vague相似性度量文献的相关性,并结合年均下载量和年均引用,对进入推荐集的文献进行推荐排序供用户选择使用。实验表明,该算法不仅利用关键词相似度,还融入了对文献有重要意义的引用数和下载量参数作为推荐参数,综合推荐效果明显。

In order to improve the efficiency of literature comprehensive recommendation,a literature recommendation method based on Vague similarity is proposed.Firstly,this method uses the literature database to retrieve the documents of interest to users as the collection of documents to be recommended.Then,according to the proposed method,the keywords in the literature are reordered,the fuzzy membership degree of keywords is calculated,and the keywords are converted to Vague values.Finally,using Vague similarity to measure the relevance of documents,and combining the average annual downloads and average annual citations,the documents entering the recommendation set are recommended and sorted for users to choose.Experiments show that the algorithm not only uses the keyword similarity,but also incorporates the references and downloads that are important to the literature as the recommendation parameters,and the comprehensive recommendation effect is obvious.