正在加载图片...
第13卷第6期 智能系统学报 Vol.13 No.6 2018年12月 CAAI Transactions on Intelligent Systems Dec.2018 D0:10.11992/tis.201802011 网络出版地址:http:/kns.cnki.net/kcms/detail/23.1538.TP.20180416.1325.008.html 融合协同过滤与用户偏好的旅游组推荐方法 陈君同,古天龙,常亮,宾辰忠,梁聪 (桂林电子科技大学广西可信软件重点实验室,广西桂林541004) 摘要:近年来,组推荐系统已经逐渐成为旅游推荐领域的研究热点之一。传统的推荐系统面临的数据稀疏性 问题在组推荐系统中同样存在。基于评分的推荐系统中,可以把组推荐系统分为对单个用户的偏好预测和对 组内成员预测结果的融合两个阶段。为提高推荐的效果,提出了一种融合协同过滤与用户偏好的旅游组推荐 方法,它考虑了用户的预测评分和组推荐结果的准确性。在协同过滤中通过加人相似性影响因子和关联性因 子进行预测评分,然后在均值策略和最小痛苦策略的基础上,提出了满意度平衡策略,该策略考虑了组内成员 的局部满意度和整体满意度。实验表明,所提出的方法提高了推荐的准确率。 关键词:组推荐:旅游推荐:数据稀疏性:协同过滤:用户偏好:均值策略:最小痛苦策略 中图分类号:TP391文献标志码:A文章编号:1673-4785(2018)06-0999-07 中文引用格式:陈君同,古天龙,常亮,等.融合协同过滤与用户偏好的旅游组推荐方法J.智能系统学报,2018,13(6): 999-1005. 英文引用格式:CHEN Juntong,GU Tianlong,CHANG Liang,etal.A tourist group recommendation method combining collabor- ative filtering and user preferences[Jl.CAAI transactions on intelligent systems,2018,13(6):999-1005. A tourist group recommendation method combining collaborative filtering and user preferences CHEN Juntong,GU Tianlong,CHANG Liang,BIN Chenzhong,LIANG Cong (Guangxi Key Laboratory of Trusted Software,Guilin University of Electronic Technology,Guilin 541004,China) Abstract:In recent years,the group recommendation system has gained much attention in the field of tourism recom- mendation.The problem of data sparsity faced by the traditional recommendation system also exists in the group recom- mendation system.In the scoring-based recommendation system,the group recommendation system can be divided into two stages:preference prediction for individual users and aggregation of the forecast results of group members.To im- prove the effect of recommendation,a tourist group recommendation approach is proposed that incorporates collaborat- ive filtering and users'preferences.It considers the accuracy of user's predictive scores and the group recommendation result.In the collaborative filtering,the predictive score is calculated by adding the similarity impact factor and the rel- evancy factor.Based on the average strategy and the least misery strategy,a satisfaction balance strategy is proposed, which considers both of the partial satisfaction and whole satisfaction of the group members.A series of conducted ex- periments show that the proposed method yields more accurate recommendations. Keywords:group recommendation;tourism recommendation;data sparsity;collaborative filtering;user's preference; average strategy;least misery strategy 随着信息技术和互联网的发展,网络正成为 息匮乏的时代走入了大数据的时代,在海量数据 人们规划旅游的重要信息来源"。人们逐渐从信 的背景下,如何快速找到对用户最有价值的信 收稿日期:2018-02-07.网络出版日期:2018-04-16. 息,显得越来越重要,推荐系统便应运而生。 基金项目:国家自然科学基金项目(61572146,U1501252,U1711263头 以往的推荐系统主要关注于单个用户,在电 广西创新驱动重大专项项目(AA17202024):广西自 然科学基金项目(2016 GXNSFDA380006). 视节目)、音乐、电影、新闻等方面取得了很好的 通信作者:宾辰忠.E-mail:binchenzhong@guet.edu.cn.. 效果,但是对于旅游领域还没有给出完善的推荐DOI: 10.11992/tis.201802011 网络出版地址: http://kns.cnki.net/kcms/detail/23.1538.TP.20180416.1325.008.html 融合协同过滤与用户偏好的旅游组推荐方法 陈君同,古天龙,常亮,宾辰忠,梁聪 (桂林电子科技大学 广西可信软件重点实验室,广西 桂林 541004) 摘 要:近年来,组推荐系统已经逐渐成为旅游推荐领域的研究热点之一。传统的推荐系统面临的数据稀疏性 问题在组推荐系统中同样存在。基于评分的推荐系统中,可以把组推荐系统分为对单个用户的偏好预测和对 组内成员预测结果的融合两个阶段。为提高推荐的效果,提出了一种融合协同过滤与用户偏好的旅游组推荐 方法,它考虑了用户的预测评分和组推荐结果的准确性。在协同过滤中通过加入相似性影响因子和关联性因 子进行预测评分,然后在均值策略和最小痛苦策略的基础上,提出了满意度平衡策略,该策略考虑了组内成员 的局部满意度和整体满意度。实验表明,所提出的方法提高了推荐的准确率。 关键词:组推荐;旅游推荐;数据稀疏性;协同过滤;用户偏好;均值策略;最小痛苦策略 中图分类号:TP391 文献标志码:A 文章编号:1673−4785(2018)06−0999−07 中文引用格式:陈君同, 古天龙, 常亮, 等. 融合协同过滤与用户偏好的旅游组推荐方法[J]. 智能系统学报, 2018, 13(6): 999–1005. 英文引用格式:CHEN Juntong, GU Tianlong, CHANG Liang, et al. A tourist group recommendation method combining collabor￾ative filtering and user preferences[J]. CAAI transactions on intelligent systems, 2018, 13(6): 999–1005. A tourist group recommendation method combining collaborative filtering and user preferences CHEN Juntong,GU Tianlong,CHANG Liang,BIN Chenzhong,LIANG Cong (Guangxi Key Laboratory of Trusted Software, Guilin University of Electronic Technology, Guilin 541004, China) Abstract: In recent years, the group recommendation system has gained much attention in the field of tourism recom￾mendation. The problem of data sparsity faced by the traditional recommendation system also exists in the group recom￾mendation system. In the scoring-based recommendation system, the group recommendation system can be divided into two stages: preference prediction for individual users and aggregation of the forecast results of group members. To im￾prove the effect of recommendation, a tourist group recommendation approach is proposed that incorporates collaborat￾ive filtering and users’ preferences. It considers the accuracy of user’s predictive scores and the group recommendation result. In the collaborative filtering, the predictive score is calculated by adding the similarity impact factor and the rel￾evancy factor. Based on the average strategy and the least misery strategy, a satisfaction balance strategy is proposed, which considers both of the partial satisfaction and whole satisfaction of the group members. A series of conducted ex￾periments show that the proposed method yields more accurate recommendations. Keywords: group recommendation; tourism recommendation; data sparsity; collaborative filtering; user’s preference; average strategy; least misery strategy 随着信息技术和互联网的发展,网络正成为 人们规划旅游的重要信息来源[1]。人们逐渐从信 息匮乏的时代走入了大数据的时代,在海量数据 的背景下,如何快速找到对用户最有价值的信 息,显得越来越重要,推荐系统便应运而生[2]。 以往的推荐系统主要关注于单个用户,在电 视节目[3] 、音乐、电影、新闻等方面取得了很好的 效果,但是对于旅游领域还没有给出完善的推荐 收稿日期:2018−02−07. 网络出版日期:2018−04−16. 基金项目:国家自然科学基金项目 (61572146,U1501252,U1711263); 广西创新驱动重大专项项目 (AA17202024);广西自 然科学基金项目 (2016GXNSFDA380006). 通信作者:宾辰忠. E-mail:binchenzhong@guet.edu.cn. 第 13 卷第 6 期 智 能 系 统 学 报 Vol.13 No.6 2018 年 12 月 CAAI Transactions on Intelligent Systems Dec. 2018
向下翻页>>
©2008-现在 cucdc.com 高等教育资讯网 版权所有