基于网络结构改进社交网络好友推荐算法研究

Based on network structure improved social network friends recommendation algorithm

随着信息技术的快速发展,人们的社交网络方式不断地改变,如何准确地给用户个性化推荐成为当前研究的热点。文中通过对基于网络结构的推荐技术深入的研究,针对该技术中随机游走模型存在的计算开销大、偏离目标及未考虑顶点重要性等不足,建立了一种基于连通性的局部随机游走重启动模型(Connectivity Local Random Walk With Restart Recommendation Algorithm,简称C-LRWR)。通过仿真实验与图模型中常见的推荐算法进行对比及测试,最后得出改进的C-LRWR算法能够提高推荐的准确率。

With the rapid development of information technology,the way of people＇s social network is constantly changing,how to accurately give users personalized recommendation has become the hot spot of the current research. This paper,through a in-depth research on a recommendation technology based on network structure,which is computational overhead,deviated from the target and does not consider the vertex importance and others in the random walk model,establishes a model based on the connectivity of the local random walk restart（ connectivity local random walk with restart recommendation algorithm,referred to as C-LRWR）. Through the simulation experiment and the comparison and test of the common recommendation algorithm in the graph model, the improved C-LRWR algorithm can improve the accuracy of the recommendation.