AI-ML Analytics: A Comprehensive Investigation on Sentimental Analysis for Social Media Forensics Textual Data

Abstract

Individuals spend a significant portion of their time on social media. It has become a platform for expression of feelings, sharing of ideas and connecting with other individuals using video and audio posts, textual data such as comments and descriptions and so on. Social media has a considerable impact on people’s daily life. In recent time, there is an enormous growth in number of people using Twitter and Instagram to share their emotions and sentiments which represents their actual feelings. In this work, we apply Machine Learning techniques on social media data to perform a comprehensive investigation to detect the risk of depression in people. Our work can help to detect various symptoms such sadness, loneliness, detachment etc. providing an insight for forensic analysts and law enforcement agencies about the person’s mental state. The experimental results show that Extra Tree Classifier performs significantly better over the other models in detecting the sentiment of people using social media data.