A NOVEL APPROACH TO DETECT SENTIMENTS BASED ON SOCIAL MEDIA USING MACHINE LEARNING

Abstract

It might be helpful to estimate a product or service's future scope by doing a sentiment analysis of it. However, it might be tedious and difficult to manually analyse a lot of papers in a short amount of time. As a result, several attempts have been made to address this issue in the literature, and various sentiment analysis approaches have been put forth. Various machine learning methods that use supervised or semi-supervised techniques are currently available. These algorithms can use a hybrid, unigram, bigram, ngram, or other suitable approach. This study makes advantage of semi-supervised learning. This study combines many approaches to create a unique model that employs a number of approaches and yields results. This fictional method produced greater result. People frequently post a lot of stuff on social media with the goal of discovering memes that capture their emotions. Because people are more inclined to convey their feelings through pictures and written descriptions, image and textual sentiment analysis (SA) is advancing at a rapid rate. Social media users are increasingly expressing themselves and sharing their experiences through photos and videos