Evaluating the Recognition of Sentiments in Text on Social Media Utilizing Machine Learning

The prevalent routine of the World Wide Web (WWW) has unveiled unique means for society to rapidly develop its moods and sentiments. It also serves as a highly content-rich platform where users may experience diverse opinions that can influence their decision-making process. When expressing opinions, both words and texts carry meanings that depend on their context and their specific lexical definitions. The interplay between these meanings creates several sentiments, i.e., positive, negative, or neutral. The text also has varied degrees of temperament. Extraction and analysis of sentiments from textual data have become a captivating field of study. Machine learning enhances this process of analyzing sentiments by applying computational algorithms to recognize temperaments expressed in textual form. Integrating these techniques to recognize feelings from a text offers promising and noteworthy outcomes. This study aims to design a model for extracting sentiments from text data, and classifying sentiments, thereby evaluating the results. The machine has been trained on a specific dataset which consists of keywords related to mental health and includes tweets for a particular year.