Recommendations in Social Network using Link Prediction Technique

Currently online social network with rapid development has become part of people's life. Based on user interest social network will change over time with different nodes and edges. Predicting new relation and missing relation between nodes in a social network can be identified using link prediction. Where new links and nodes can be identified with attribute information. Machine learning approaches are used with a set of features to increase the performance using supervised learning setup. Predicting the probability of connection between nodes for Twitch dataset, collected from SNAP and the main objective is to predict the probability of connection between nodes for personalized recommendations of game streamers using supervised machine learning by training model and performance of the model is evaluated using prediction performance metrics.