融合评论标签的个性化推荐算法

发布时间：2018-11-14 09:37

【摘要】：传统的推荐算法大都从评论中挖掘用户兴趣或产品特征,然而由于评论形式自由,规则性差,导致从评论中获取有效信息较困难,推荐结果不理想。在电子商务等领域,评论标签作为一种新的评论方式已经被广泛使用。与评论相比,评论标签具有规则性强,信息密度大等特点,因此提出了一种融合评论标签的推荐算法。该算法从评论标签中挖掘用户对产品特征的观点,并利用其构建用户兴趣模型和产品特征模型,然后向用户推荐在他们感兴趣的特征上有较高评价的产品。与传统推荐算法进行对比,实验结果表明,融合评论标签的算法能有效地提高用户的覆盖率,并提升推荐算法的准确性。
[Abstract]:Most of the traditional recommendation algorithms mine user interest or product features from comments. However, due to the freedom of comment form and poor rules, it is difficult to obtain effective information from comments, and the recommended results are not ideal. Comment labels have been widely used as a new way to comment in e-commerce. Compared with comments, comment tags have the characteristics of strong regularity and high information density. Therefore, a recommendation algorithm is proposed to fuse comment tags. In this algorithm, users' views on product features are mined from comment labels, and then user interest models and product feature models are constructed, and then products with high evaluation on features of interest to users are recommended to users. Compared with the traditional recommendation algorithm, the experimental results show that the algorithm can effectively improve the user coverage and improve the accuracy of the recommendation algorithm.
【作者单位】：福州大学数学与计算机科学学院;
【基金】：国家自然科学基金,No.61300105 教育部博士点基金联合资助项目,No.2012351410010 福建省科技重大专项,No.2013H6012 福州市科技计划项目,Nos.2012-G-113,2013-PT-45~~
【分类号】：TP391.3

【相似文献】