中文摘要：

推荐系统能够有效解决用户面临的信息过载问题。移动环境下,用户对信息的需求随着情境变化而不断变化,而且不同情境下,具有共同兴趣的用户形成了不同的群组。因此,文章提出了基于用户共同兴趣的情境感知信息推荐方法。首先设计了基于散列的共同兴趣挖掘算法获取用户群组的共同兴趣特征;然后根据共同兴趣关联,计算用户和项目的关联度,并通过共同兴趣协同预测用户对项目的偏爱程度;最后综合以上两个因素进行情境感知的信息推荐。实验表明,提出的情境感知信息推荐方法在推荐质量和效率上均优于现有方法。

英文摘要：

Recommendation system can effectively solve the information overload problem faced by users. Under the mobile environment, the information demands of users change with the context and users with common interests form different groups in different context. Therefore, this paper proposes context awareness information recommendation approach based on common interests of users. The paper first designs common interests mining algorithm based on hashing to retrieve charaeteristies of common interests of user groups. Then, the paper calculates the relevancy degree between users and items according to common interest, and predicts users＇ preference to items through collaborative filtering based on common interests. The experimental results indicate that the proposed context awareness information recommendation approach outperforms state-of-the-art approaches in both reeommendation quality and efficiency.