Abstract:

Aiming at the problem of how to provide efficient and personalized service for passengers of high-speed railway stations in real time, this article introduces context information into the service recommendation process, and analyzes user context information according to time, space, user, and application in four aspects. The growth algorithm enables it to support the mining of two-dimensional data items and obtain the association rules between context and service functions. Through the similarity calculation with the current context information, a personalized service set available in the current application scenario is constructed. Experiments show that this method can significantly improve the accuracy and hit rate of service matching, and help to improve passenger service satisfaction.

Key words: context-aware, association rules, mobile application, similarity, railway passenger service

摘要：

针对如何为高铁站旅客即时提供高效的个性化服务问题，将上下文信息引入到服务推荐过程中，对用户上下文信息按照时间、空间、用户、应用四个方面进行数据分析处理，采用改进FP-Growth算法，使之支持对二维数据项的挖掘，获取上下文与服务功能之间的关联规则。通过与当前上下文信息的相似度计算，构造出当前应用场景下可用的个性化服务集合。实验证明该方法可提高服务匹配的准确率和命中率，有助于提升旅客服务满意度。

关键词: 上下文感知, 关联规则, 移动应用, 相似度, 铁路客运服务