A Detailed Survey on Sentimental Analysis on Social Media

Abstract

Public opinion or point of view matters a lot as it plays vital role in shaping country’s image all around the world, may it be on certain products, decisions or their own feelings/expressions. At this time, people are more expressive on social media, more rather than in person, they express or share their daily tasks their feelings, opinions some good or bad incidents, and many more on social media, and we can collect this data and process into classifiable form. The data can be collected through API and then transformed into classifiable form. To classify data, we will be foreseeing three algorithms which are Naïve Bayes, maximum entropy, and lexicon-based. The main reason of gathering this data is to process and generate positive and negative notions of certain demography or specific person. This data can be useful in many ways such as to check if a person is in good mental health condition or suffering through depression through his/her positive negative posts on social media. Multi-national companies can collect insights on their products based on peoples review and opinion on social media. MNC and e-commerce stores can also product recommendation system based on those reviews. In short, sentiment analysis plays a very vital role in our economy as well as health care.