Tamil News filtering with Aspect-based Sentiment Analysis for Children

Abstract

" Reading Newspapers is the best way for kids to learn a language and get information about what is going on in our local communities and globally. Because of the exponential growth of the Internet and information technology, reading online news feeds via browsing the Internet has become very common. However, the central problem of the online newspaper is, it carries negative and violent articles. In this research project, the author attempts to filter out these negative and unsuitable articles for children’s reading. By doing so kids only read positively written news articles that ensure their positive emotional state. The Tamil News filtering with Aspect-based Sentiment Analysis for Children system will be a novel system which has not been created ever before and compared to the existing similar sentiment analysis systems this system outperforms in terms of accuracy. This system built with Python as its primary language and BERT model is developed as the core model. This system aims to collect live news articles and predict the sentiment polarity on those news articles using RSS link."