基于移动互联网的文献个性化推荐系统

Mobile Internet based personalized literature recommendation system

为了满足科技工作者对文献获取和阅读的便捷性、高效性及个性化的需求,提出了一种基于移动互联网的文献个性化推荐(MIPLR)系统,并阐述了系统的体系架构以及各个模块的功能及实现。通过实现基于接受者请求的推送模型,解决了大量移动用户内容自动获取的问题;构建科技文献领域的垂直搜索引擎,解决了文献资源获取的问题;采用向量空间模型(VSM)表示文献资源与用户模型以及使用用户阅读行为反馈机制动态更新用户模型,来推荐准确的文献资源给用户。实验结果表明,该系统有效地提高了用户文献检索效率和推送结果的准确率,可以持续、自动地为用户提供符合其兴趣爱好的文献资源,为用户节约了大量的时间与精力。

To meet the needs of reading literature conveniently, efficiency and personally, a Mobile Interuet based Personalized Literature Recommendation （MIPLR） system was proposed. The architecture of the system as well as the function of each module was expounded. To support retrieving content automatically by large volume of users, a receiver intent based sender push model was implemented. A vertical search engine for literature was constructed for literature acquisition. For recommending suitable literature resources to user, this system used Vector Space Model （VSM） to represent literature resources and user model and dynamically updated user model according to user behavior. The experimental results show that this system can improve user literature retrieval efficiency and increase the accuracy of the literature recommendation. MIPLR can automatically provide users with literature resources and saves users＇ immense amounts of time.