个性化旅游景点推荐研究

发布时间：2019-01-10 11:33

【摘要】：随着移动互联网的广泛普及和旅游者对旅游服务品质的要求不断加深,在线旅游、移动旅游等服务也逐渐兴起,在线旅游景点的个性化推荐逐渐成为个性化推荐技术领域的一个应用和研究热点。面对庞大、复杂的旅游数据,旅游者对于旅游景点的个性化服务的需求也越来越强烈,研究高效、准确的个性化旅游景点推荐系统具有很好的应用价值。本文针对个性化旅游景点推荐的应用需求,借助于社交网络与贝叶斯网络,充分挖掘用户与景点之间的匹配度进行个性推荐。本文主要研究工作如下:(1)提出基于社交网络的个性化旅游景点推荐算法。为了提高旅游景点推荐的准确率,解决新用户冷启动问题,该算法将社交网络因子加入到旅游景点推荐中,充分挖掘用户间的社会化网络关系。该算法处理过程如下:首先,采用耦合双向聚类算法对用户进行聚类处理;然后,使用DBSCAN算法对景点聚类;最后将两个稳定的用户集合和景点集合应用在个性化推荐算法上,预测用户下一个即将去的景点。在数据集上将该算法与一些传统的算法进行了对比实验,实验结果表明本文提出的算法具有较高的推荐准确率。(2)为了量化旅游景点推荐,本文提出基于贝叶斯网络学习的个性化旅游景点推荐算法。该算法为了解决新用户与新景点的问题,综合使用了用户的人口统计学信息,用户-景点评分信息以及景点属性。具体地,该算法首先使用传统的协同过滤算法处理用户属性相似度与用户行为相似度,使用基于内容的算法处理景点间关系;然后,使用贝叶斯概率模型计算出用户访问每个景点的概率;最后,将此算法在携程网数据集上与传统的算法进行实验验证,结果表明该算法在处理新用户和新景点问题上具有更好的性能。
[Abstract]:With the widespread popularity of the mobile Internet and the deepening requirements of tourists for the quality of tourism services, online tourism, mobile tourism and other services are also gradually rising. Personalized recommendation of online tourist attractions has gradually become an application and research hotspot in the field of personalized recommendation technology. In the face of the huge and complicated tourism data, tourists' demand for personalized tourist attractions service is becoming more and more intense. Therefore, the study of highly efficient and accurate personalized recommendation system of tourist attractions has a good application value. This paper aims at the application demand of personalized tourist attraction recommendation, with the help of social network and Bayesian network, fully excavates the matching degree between user and scenic spot to carry on personality recommendation. The main research work of this paper is as follows: (1) A social-networking based personalized recommendation algorithm for tourist attractions is proposed. In order to improve the accuracy of recommendation of tourist attractions and solve the cold start problem of new users, the algorithm adds the social network factor to the recommendation of tourist attractions, and fully excavates the social network relationship among users. The processing process of the algorithm is as follows: firstly, the coupled bidirectional clustering algorithm is used to cluster the users; then, the DBSCAN algorithm is used to cluster the scenic spots. Finally, two stable user sets and attraction sets are applied to the personalized recommendation algorithm to predict the next destination that users will go to. The experimental results show that the proposed algorithm has a high recommendation accuracy. (2) in order to quantify the recommendation of tourist attractions, the proposed algorithm is compared with some traditional algorithms. This paper proposes a personalized recommendation algorithm for tourist attractions based on Bayesian network learning. In order to solve the problem of new users and new scenic spots, the algorithm uses the demographic information of users, the information of user-scenic spot score and the properties of scenic spots. Firstly, the traditional collaborative filtering algorithm is used to deal with the similarity between user attributes and user behavior, and the content-based algorithm is used to deal with the relationship between scenic spots. Then, the Bayesian probability model is used to calculate the probability of user visiting each scenic spot. Finally, the algorithm is tested on Ctrip data set with the traditional algorithm. The results show that the algorithm has better performance in dealing with new users and new scenic spots.
【学位授予单位】：天津理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.3

【相似文献】