弱连接边缘独立判别社交网络社区快速生成树推荐

针对传统社交网络社区推荐算法精度不高且计算复杂度过高的问题,提出一种弱连接边缘独立判别社交网络社区快速生成树检测算法,在提高社区推荐精度同时,降低算法计算复杂度。首先,结合社交网络社区推荐特点,设计基于边缘重量分配节点相似性的最大生成树算法,实现对社交网络社区的有效检测；其次,针对所提算法,存在弱连接边缘重复添加、删除,浪费计算资源的问题,提出弱连接边缘独立判别的快速生成树检测算法,进一步提高算法计算效率；最后,通过在标准测试数据库中的实验对比,验证了所提算法的有效性。

WERSTA: Weakly connected edge independent discriminant based rapid spanning tree algorithm for social network community detection

According to the problem of low accuracy and high computational complexity in traditional social network community recommendation algorithm, here proposed the weakly connected edge independent discriminant based rapid spanning tree algorithm for social network community detection, which could improve the community recommendation accuracy, and also reduced the computational complexity of the algorithm. Firstly, combined with the social network community characteristics, here designed the edge weight distribution node similar based maximum spanning tree algorithm, which realize the effective detection of the social network community; Secondly, in order to solve the problem of proposed algorithm in the existence of weak connection that adding and deleting the weak edge could waste a lot of computing resources, here proposed the rapid spanning tree algorithm with weakly connected edge independent discrimination detection, which could further improve the efficiency of algorithm; Finally, through experiments in the standard test database, verified the effectiveness of the proposed algorithm.