随着Web 2.0理念的逐渐深入及社会网络的迅猛发展，电子商务逐渐向社会化商务转变，电子商务推荐技术面临着新的机会与挑战。如何从社会网络蕴藏的海量数据中提取出有价值的情境信息，融合到传统推荐技术中，以提高推荐系统的性能和效率、并提升电子商务企业绩效，成为下一代推荐系统的研究热点。为此，本课题将对融合情境信息的多目标社会化商务推荐系统的关键理论和技术展开研究。首先，研究基于社会网络挖掘的情境信息获取，重点研究社区结构和社区意见两类与推荐密切相关的情境信息的挖掘与演化；其次，研究融合情境信息的电子商务推荐模型、算法及演化机制，解决情境推荐生成问题；再次，从企业角度出发，分析商品成本、销量和效益之间的关系，从而研究推荐结果优化策略，以增强推荐系统效果。在理论研究的基础上，设计并实现原型系统作为理论研究验证平台，并将原型系统应用到电子商务企业的案例研究中，通过案例研究为电子商务企业提升绩效服务。

As the rapid development of Web 2.0 and social network, e-commerce is gradually turning to social commerce, which leads to bran-new opportunities and challenges to recommender technologies. How to combine contextual information mined from social networks with traditional recommender techniques has become a hot research topic, so that the effectiveness and efficiency of recommender systems as well as the performance of e-business enterprises will be enhanced. The project will conduct research on key theory and technology of multi-targeting recommender systems incorporating contextual information in social commerce. Firstly, the contextual information retrieval technology based on social network mining will be investigated. We will focus on the mining and evolution of two types contextual information that are closely related to recommender systems; that are, communities and their opinion. Secondly, to solve the contextual recommendation generation problem, recommendation models, algorithms and their evolution mechanisms for incorporating contextual information will be studied. Thirdly, from the perspective of e-commerce enterprises, the relationship among product costs, sales volume and benefits will be analyzed, and thus the optimization strategy for recommendation output will be explored. Finally, a prototype system will be designed and implemented. Simulation experiments on open data sets will be conducted to verify the results of our research. The prototype system will also be applied to e-business companies for case study, which in turn justify the value of our research and helps to improve the company performance.