Nowadays, location-based social network (LBSN) has become one of the most popular applications with the rapid development of mobile Internet. However, due to the spatial and real-time properties, mobile service recommendation under LBSN environment faces too many challenges especially data sparsity problem.To tackle these challenges, a recommendation framework is proposed in this paper which has four layers defined as data collection layer, user profile modeling layer, information processing layer and recommendation feedback layer, respectively. Furthermore, the ISC-CF algorithm is implemented to integrate users’ interest profile, social influence and current location context to effectively overcome the data sparsity problem. Thus, the social influence is quantified by a modified measure way. Finally, a dynamic and personalized adjustment algorithm is built by using the users’ profile tracking and the current location context. The experiment results show that the algorithm proposed in this paper has significantly superior performance compared with the other baseline recommendation methods in both hometown area and out-of-town area.

中文翻译：

---

LBSN环境下的用户画像感知服务协同推荐算法

如今，随着移动互联网的飞速发展，基于位置的社交网络（LBSN）已成为最流行的应用之一。然而，由于其空间和实时性，LBSN环境下的移动服务推荐面临着太多的挑战，尤其是数据稀疏问题。为了应对这些挑战，本文提出了一个推荐框架，它有四个层定义为数据收集层，用户画像建模层、信息处理层和推荐反馈层。此外，实施 ISC-CF 算法以整合用户的兴趣概况、社会影响力和当前位置上下文，以有效克服数据稀疏问题。因此，社会影响力是通过改进的测量方式来量化的。最后，利用用户画像跟踪和当前位置上下文构建动态个性化调整算法。实验结果表明，与其他基线推荐方法相比，本文提出的算法无论是在家乡地区还是在外地地区，都具有显着的优越性能。