Human Sentiment Analysis on Social Media through Naïve Bayes Classifier

Abstract

350 DOI: 10.37398/JSR.2022.660137 Abstract: Deciphering feelings and thoughts from a succession of words is one of the most complex and demanding undertakings. Recognizing sentiments and emotions is one of the most effective ways of expressing feelings and sentiments by writing text. It requires more interest from researchers in advancement than face or voice-based systems. Text based emotion analysis has sparked the attention of many individual researchers to continue their research into distinguishing unique emotions from natural language. The emotion recognition from text field is used in a range of applications, such as recommendation systems, cultural content services that recommend music based on a user's current emotional state, mood tracking, emotion retrieval from suicide notes, capturing emotions in multimedia tagging, detecting objectionable phrases in chats, and so on. In today's informationrich culture, smart sociotechnical systems are gaining traction, with various technologies being employed to gather data from such systems and analyze that data for useful insights into our daily activities. Recent advancements in health monitoring and communications technologies, among other noteworthy achievements, have helped sentiment identification. The trend in artificial intelligence (AI) research in recent years has been to incorporate AI techniques into daily living objects. It is well understood that AI systems will be beneficial to the majority of humans. Emotions are a collection of mental states brought on by a variety of feelings, ideas, and behaviors. People continually communicate emotional cues during the communication process; emotional awareness is vital in human interaction and in many facets of daily life. The seven emotional states (disgust, neutral, happy, sad, angry, astonished, and bored) are extensively described in this study in order to incorporate user text emotions through social media platforms using Correlation based Naive Bayes Classifier and achieve an accuracy rate of 99.99%.