基于RippleNet的实体加权新闻推荐  

Entity Weighted News Recommendation Based on RippleNet

作　　者：黄华靖 韩梅 刘康民 刘翥 范永全 HUANG Huajing;HAN Mei;LIU Kangmin;LIU Zhu;FAN Yongquan(School of Computer and Software Engineering,Xihua University,Chengdu 610039 China;Chengdu Siwei Power Electronic Technology Co.,Ltd.,Chengdu 610036 China)

机构地区：[1]西华大学计算机与软件工程学院,四川成都610039 [2]成都四威功率电子科技有限公司,四川成都610036

出　　处：《西华大学学报(自然科学版)》2025年第2期79-86,共8页Journal of Xihua University:Natural Science Edition

基　　金：国家自然科学基金项目(61872298、61802316、61902324);四川省科技创新苗子工程(2015075);四川省科技厅项目(2023YFQ0044);西华大学研究生科创竞赛项目(YK20240148)。

摘　　要：推荐系统一直是信息检索研究的热点。相比于电影推荐、旅游推荐等项目推荐,新闻推荐具有新闻文章批量大、时效性强的特点,从而对算法有着更高的要求。文章在RippleNet模型的基础上,引入实体入度的概念,通过把多边实体进行加权,突出个别实体的重要程度,以提高新闻推荐的精度。在新闻数据集的性能验证结果表明,该模型相比于其他基线模型,整体精度提升了1.7%。Recommendation system has always been a hot issue in information retrieval research.Compared with other items such as movie recommendation and travel recommendation,news recommendation has the characteristics of large batch of news articles and strong timeliness,which has higher requirements on the algorithm.Based on the RippleNet model,this paper introduces the concept of entity entry degree,which highlights the importance of individual entities by weighting multilateral entities,thus improving the accuracy of news recommendation.The performance validation on the news dataset shows that the overall accuracy of this model is improved by 1.7%compared with other baseline models.

关 键 词：推荐系统 新闻推荐 RippleNet 实体入度 

分 类 号：TP391.3[自动化与计算机技术—计算机应用技术] G210.7[自动化与计算机技术—计算机科学与技术]