Abstract: The link prediction problem in scale free networks is studied. Based on the man-made and real social network, the general processes of static and dynamic prediction are given respectively. The accuracy of several similarity methods is verified, and therefore, the method of using similarity to predict links is proved to be feasible. The pridiction algorithms are recommendied according to the effectiveness of pridiction results.

Key words: complex network, information retrieval, scale-free, link prediction, topological structure, similarity

摘要： 研究无标度网络中的链路预测问题。针对人造网络和实际社会网络，分别介绍静态和动态2种链路预测的实现过程，探究利用相似性进行链路预测的可行性，并验证多种相似度计算方法的准确性。对预测结果进行有效性分析，同时根据不同网络特性给出相应的预测算法。

关键词: 复杂网络, 信息检索, 无标度, 链路预测, 拓扑结构, 相似性

CLC Number: 

• N945