GOOGLE PLAY STORE USERS COMMENT REVIEW CLASSIFICATION USING SVM CLASSIFIER AND RANDOM FOREST

Abstract

In today's digital age, social media stands as a dynamic arena where individuals freely express their thoughts and opinions, from succinct tweets on Twitter to expansive narratives on platforms like Facebook and Instagram. However, amidst this vast sea of user-generated content, a glaring void persists a definitive rating system capable of distilling the nuanced sentiments embedded within these diverse commentaries. This study thus emerges as a pioneering endeavor, poised to bridge this crucial gap in sentiment analysis. Leveraging the transformative potential of the Word2vec methodology in the preprocessing phase, researchers embark on a comprehensive journey to classify comments on a meticulous 1-5 rating scale, thereby unraveling the multifaceted spectrum of sentiments encapsulated within them. Complementing this groundbreaking approach, the Random Forest classification model is harnessed to bolster the analytical prowess of the study. The resultant accuracy score of 60.4% stands as a testament to the study's significant strides towards achieving a deeper understanding of comment sentiment in the realm of social media. However, this is merely the inception of a promising trajectory; the study's findings hold the promise of not only refining sentiment analysis techniques but also revolutionizing diverse sectors, from market research to product development. With this study, the narrative of sentiment analysis transcends the confines of academia, beckoning forth a new era of nuanced comprehension and meaningful engagement within the sphere of social media commentary. As the study concludes, it leaves behind a compelling call to action, inviting further exploration and innovation in this dynamic field.