Corpus Usage for Sentiment Analysis of a Hashtag Twitter

Social media (Facebook, Instagram, Twitter, etc.) nowadays can be used for analyzing the objects, e.g. political views, products, services, etc. To understand the performance of an object, the sentiment analysis has been widely used to get the review from the consumers or users (positive or negative responses). Today, in big date era, which is a component of industry 4.0, many corpora are available and can be accessed freely. A corpus can be utilized to train the model through some methods. In this paper a Naïve-Bayes classifier was used to train a corpus from natural language toolkits (NLTK) corpora. As a case study, sentiment analysis for the sample movie “Avengers” was done from the twitter hashtag #avengersendgame. The paper also proposed the usage of a particular corpus to other different language implementations, e.g. for Indonesian language. Through the use of Tweepy and Pandas some Twitter tweets were retrieved and classified after pre-processing. The results showed the capability of the Naïve-Bayes classifier both for English and Indonesian language.