Decoding Redditors Emotional Cues: A Naïve Bayes Theorem Unveils the Hidden Stress of Subscribers Through Sentiment Analysis

Abstract

Social media is gradually replacing in-person interactions in communication including sharing information and opinions, describing experiences, and updating others on daily activities. Virtual communication is replacing real-world communication in certain ways. Reddit is a social news website and forum where users may vote on, comment on, and share items, it is a conversation site. It is divided into subreddits, or communities centered around specialty subjects such as technology, politics, or more. The massive number of posts, content, comments, and discussions on different specific topics provides a chance for behavioral research and sentiment analysis on humans. The Redditor's behavior and the emotional cue represent an enormously expansive and mostly unknown research field. Stress detection is one of these fields. This paper's primary goal is to conduct the sentiment analysis using the Naïve Bayes approach to predict the stress level of Redditors based on the posts and comments they shared in the stress community on their social platforms. The secondary goal is to help manage stress and reduce negative emotions; we recommend exploring various coping strategies. To sum up, the detection rate of Naïve Bayes prediction is quite good. The results of this research are beneficial to one's well-being.