Extracting Human Sentiments to Perform Link Prediction using Machine Learning

Abstract

Communication Networks like Social Network are complex and requires lengthy computation. It is a challenging task to preprocess the network by removing the redundancies. After studying almost every related approach dealing with these complex network, we have found that none of the approaches is stable and durable. Stability means the approach should work on any network either a static or dynamic, and durability means approach should be consistent. We have designed a model that works on sentiment-based link prediction. It is hard to customize the model on a single platform. Thus, we divide our model into three phases. We then separately execute these phases. In the first phase, we developed US-Frequency approach to study and analyze the common sentiment. In the second phase, we developed Shabaz-Urvashi Link Prediction to predict the future links, and In the third phase, we validate our model with the help of case studies newline