Automatic Sentiment Analysis System for Myanmar News

Abstract

Sentiment analysis and opinion mining mainly emphasize on opinions which express or describe positive or negative sentiments. Sentiment analysis is mostly concerned with NLP. The inception and rapid development of the field concerned with the social media on the Web, e.g., post, reviews, forum discussions, blogs, microblogs, Twitter, and social networks. There is a little research in Myanmar language. This system implements sentiment analysis system for Myanmar news using well-known machine learning approach such as Support vector machine. Feature extraction and selection are also need for sentiment analysis to improve accuracy. This system uses N gram and TF-IDF method for feature selection and extraction. News corpus is used for training and testing. This system uses News data set from different web pages in Facebook social media.