中文摘要：

针对大规模社交网络应用中检索结果过于庞大复杂的问题，将个性化推荐与可视化相结合，用于在大量数据中找到用户感兴趣的信息。在开拓网络缩放算法的基础上，提出关键信息显示算法，能够区别显示社交网络关系图中用户相对重要的信息和次要信息，增强关联度较高数据的显示效果。将带权值的力导向布局算法应用于用户关系聚类中，通过在二维显示空间中合理安排节点布局，达到减少用户认知负担和个性化推荐的目的。设计并实现个性化推荐的可视化工具HRVis，在Movielens数据集上进行测试，结果表明，HRVis能够强调显示具有良好社会关系的重要用户以及与用户相似的关联用户，获得较好的可视推荐效果。

英文摘要：

Aiming at the problem that the search results are too huge and complex in the application of large-scale social networks, the method of combining personalized recommendation with visualization is proposed to find useful information from mass data. This paper proposes a key information display algorithm on the basis of Pathfinder Networks（PFNET） algorithm to display key information by emphasizing important records with high degree of associations. Additionally, an improved weighted force-directed algorithm is applied to cluster user relations to improve displaying layout for the purpose of facilitating users＇ cognition and achieving personalized recom- mendation. It designs and implements a visualization of personalized recommendation tool HRVis. The Movielens datasets shows that the HRVis can emphasize important users which have good social relations and associated users which are similar to the users, and it has good visual recommendation effect.