An Improvised Method for Anomaly Detection in social media using Deep Learning

Abstract

Recently, social media has arisen not only as a personal communication media, but also, as a media to communicate opinions about products and services or even political and general events among its users. Due to its widespread and popularity, there is a massive amount of user reviews or opinions produced and shared daily. Twitter is one of the most widely used social media micro blogging sites. In this paper, a deep learning-based approach is developed to detect the anomalies in social media using text mining. The emotional classification is considered as a part of the model that classifies emotional anomalies present in the text. Classification of such text is conducted via proper training and testing of the classifier.