Performance evaluation of similarity-based link prediction schemes for social network

Social network is a platform where people can share their information, make new connections, and explore the information about the different events occurring in society. In recent year, prediction for new link in the social network has been attracted many researchers. Link Prediction in social network refers to finding new connections that will occur between people in the future. It finds the absence and presence of edge in a social network. In this paper, we have explored well known similarity based link prediction algorithms and discuss their performance in terms of accuracy, precision, and recall. The experimental results show that common neighbours and Jaccard coefficient based algorithms perform better than other algorithms in term of accuracy.