Abstract
WWW has rapidly developed into a cutting-edge platform for people to voice their views and ideas on various subjects, trends, and problems. The contents created by individuals in different platforms including online forums, chatting platforms, blogs, etc. serve an important role in decision-making. Advertising, opinion polls, online surveys, market forecast, corporate information, social media discussions, etc. are the primary sources of content creation. Sentiment analysis deals with the issue of extracting sentiments from text data and classifying the author's viewpoint on a specific entity into no more than three predetermined categories: positive, negative, and neutral. Still now there exists any article which elaborates step by step procedure using Python in a user friendly manner. In this regard, in this article, we describe the step by step sentiment analysis procedure to categorize Twitter's highly unstructured data (as a case study) in a very user friendly approach using Python and also we compare the performance of different machine learning techniques discussed here. The conclusion of this evaluated study reflects the effective and usefulness of different machine learning techniques for sentiment analysis.
Recommended Citation
Bhattacharya, Sayantan; Hasan, Mahmadul; Ali, Md Rahmat; Ghayas, Fahad; Das, Chandra; and Bose, Shilpi
(2024)
"Comparative Study on Machine Learning Techniques based Sentiment- Analysis of Textual Documents from Twitter,"
American Journal of Advanced Computing (AJAC): Vol. 2:
Iss.
4, Article 5.
Available at:
https://research.smartsociety.org/ajac/vol2/iss4/5