Detecting Sentiment from Bangla Text using Machine Learning Technique and Feature Analysis

Abstract

Sentiment Analysis is an ongoing field of research in text mining, especially in the area of Bangla language as there are few research works done in this particular sector. In general, sentiment classification means the analysis to determine the expression of a speaker whether he or she holds a positive or negative opinion to a specific subject on a given text. It is consider the information in a text at the document, and extract unique feature/aspect level whether the given query holds an expression of positive, negative or neutral. In this process of retrieving information of a document the author use Tf.Idf (term frequency–inverse document frequency) to come out a better solution and give more accurate result by extracting different feature of a positive, negative or neutral word of sentiment analysis in particular view of Bangla text. The author calculate the total positivity, negativity of sentence or document with respect to total sense. Sufficient example and experiment are presented to describe the feature extraction of sentiment that it‘s found in this methodology. General Terms Machine learning, Bangla text analysis