| 摘 要: 推荐系统是通过分析已知信息和用户偏好，在用户选择物品或服务时，向用户提供帮助和建议的系统。 但是目前大部分推荐系统都是基于用户评价或评分信息向用户推荐购物、电影等电子商务服务，基于用户轨迹数据进行 用户兴趣区域推荐的研究十分罕见。用户的轨迹数据蕴含了用户的偏好，不同的轨迹反映不同的用户特性。所以提出一 种从用户轨迹数据中挖掘最大频繁项集，并将最大频繁项集用于计算用户相似性和偏好的推荐方法。该推荐方法还综合 考虑了相似用户访问次数、置信度和用户住宅信息等可能会影响推荐质量的因素。将提出的方法和基于协同过滤的推荐 方法、基于关联规则的推荐方法进行比较，结果显示本文提出方法的效果较好。 |
| 关键词: 轨迹数据挖掘；区域推荐；相似用户；频繁项集 |
|
中图分类号: TP391
文献标识码: A
|
| 基金项目: 国家自然科学基金项目(61966036、61662086)；云南省创新团队项目(2018HC019). |
|
| User Interest Region Recommendation Based on User Trajectory Data |
|
LONG Yurong,WANG Lizhen,CHEN Hongmei
|
( School of Information Science and Engineering, Yunnan University, Kunming 650504, China)
|
| Abstract: A recommendation system is a system which provides help and advice to users by analyzing the existing information and users' preferences when users choose goods or services.However,most recommendation systems recommend shopping,movies and other e-commerce services to users based on user evaluation or scoring information.It is very rare to conduct research on user interest region recommendation based on user trajectory data.User's trajectory data contains user’s different preference,reflecting different user characteristics.Therefore,it is necessary to make recommendations based on user trajectory data.This paper presents an interest region recommendation method,which calculates user similarity and preference by mining maximum frequent itemsets from user trajectory data.We take into account three factors,including numbers of similar user visits,confidence and user residence information.In the paper,the proposed method is compared with the recommendation algorithm based on collaborative filtering and the recommendation algorithm based on association rules,and the results show that the proposed method is effective. |
| Keywords: trajectory data mining;region recommendation;similar users;frequent itemset |
