Behavioral Analysis of Depressed Sentimental Over Twitter: Based on Supervised Machine Learning Approach

Abstract

Microblogging is an especially prevalent broadcast medium amidst the net fraternity currently. Social networking sites and web access have become pervasive and a basic element, they have emerged as a major platform for people to share their emotions, opinions, and reviews. The essence of the social content posted, the expressed feelings hidden in the exposed content can be greatly analyzed by sentimental analysis. Consequently, the possibility and growth of cybercrime have enhanced also. To reduce these case, we propose a method during this paper “Twitter - surveillance based on supervised machine learning approach” that can discover different harmful activities, depressed sentimental content and cyber crimes like commit suicide, physical assault, the misuse of photographs or recording of a pornographic, erotic or awkward nature, abusive trolling, blackmail, spam, fraud etc. from the social network sites. Social networking sites can be preventing such activities from providing timely notifications through helpline notification. The goal of this study is to classify twitter data with more accuracy and reliableness into sentiments by exploiting different-different supervised machine learning classifiers.