A comparative study of opinion mining utilizing twitter data for depression detection

Abstract

The use of social media has a significant impact on human life. People use this platform for a variety of purposes, including entertainment, information retrieval, news, commerce, and many others. It is possible to identify whether data is good, negative, or neutral using Sentiment Analysis, which is a type of natural language processing technology. Sentiments are acquired from a variety of sources, including conversations, blogs, tweets, and other social media. The tremendous rise of social media has aided in the development of this topic as a research hotspot in recent years. Nowadays, every corporate organization seeks to understand the public's perception of their products by conducting a variety of surveys in order to improve their operations in the future. This type of platform allows ordinary people to express their emotions on many themes with multiple users through the usage of social media networks. Multiple audio, video, and text formats are available for us to share our daily thoughts, comments on current events, as well as our political views on a variety of topics. Although Social Media provides a variety of connection options, text mode remains the most effective method of communicating our thoughts to others. People can express their feelings about a variety of subjects through the use of text messages known as "tweets." As a result, data from the Twitter application has been selected for analysis. In this work, we will investigate a complete study of sentiments from Twitter, where people express their feelings in the form of tweets and provide helpful information gleaned from the data mining process. It is our intention to look at a comprehensive technique that has been implemented in Sentiment Analysis where Naïve Bayes Classification technique has been adopted to train and test the data to predict depression level of each user. Depression score of every users will be calculated and the final opinion will be delivered based on the score. This concept will be beneficial for medical ground.