针对赛博空间中的社会舆情数据链路预测的问题,提出将网络节点活跃度作为一个重要的指标整合到预测算法中,以提高链路预测的准确性和科学性。首先定义了节点活跃度并构建三个用于表示节点活跃度的函数,然后以实际数据为样本进行实例验证,最后通过与三个主流算法在使用三个函数前后的准确度进行比较,链路预测的准确度有明显提高。
This paper derives a new link prediction algorithm that synthesizes network node active degree to improve the accuracy. In this.study, a definition of node active degree and three functions was given. Real data from Tianya. cn are used to verify this algorithm and the result of simulation, which is compared to three well-known algorithms. The result shows the accuracy is significantly enhanced, it argues for the importance of applying node link degree on the link prediction algorithms.