基于认知模型的社会关系预测算法

A SOCIAL RELATIONSHIP PREDICTION ALGORITHM BASED ON COGNITIVE MODEL

针对低采样率下社会网络中传统的关系预测方法精度较低的问题,提出一种基于认知模型的社会关系预测算法。该方法利用了单个节点对整个网络的认知能力,部分随机采样获得采样节点对社会网络中节点间关系的认知信息,然后根据认知信息预测出未采样节点的社会关系,实现了低采样率下所有节点间社会关系的预测。为了分析算法性能,在不同网络中用该算法与传统方法进行多组对比实验,结果表明该算法在低采样率下提高了预测精度、降低了预测时间。

A new social relationship prediction algorithm based on cognitive model was developed to solve the low accuracy problem of traditional prediction method in social networks under low sampling rate. On the basis of the cognitive ability of single node on whole network,this new method acquires the cognitive information of the sampled nodes on the relationship between nodes in social networks by sampling partly and randomly,then predicts the social relationship of those nodes haven＇t been sampled,and realises the prediction of the social relationship among all the nodes under low sampling rate. In order to analyse the performance of this algorithm,groups of comparative experiments between this algorithm and traditional methods were conducted in different networks. Results showed that this algorithm improved the prediction accuracy and reduced prediction time under low sampling rate.