期刊名称：International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
印刷版ISSN：2278-1323
出版年度：2017
卷号：6
期号：6
页码：953-956
出版社：Shri Pannalal Research Institute of Technolgy
摘要：Link prediction is a significant factor of any social networking site analysis. Social network sites like Facebook and LinkedIn use link prediction as a feature. In this paper, we propose a technique based on node similarity measure and hierarchical clustering. We first calculate the similarity score using Jaccard’s coefficient and then assign the score to all pairs of nodes and apply an Artificial Neural Network to the system. This technique improves the accuracy of Link Prediction and is backed by results that prove the proposed model outperforms other methods when both precision and recall of the prediction results are considered.
关键词：Hierarchical Clustering; Jaccard’s Coefficient; Link prediction; Neural Networks; Similarity Score.
