The Sentiment Analysis in Opinion Mining of Web News Comments

Qing Du,Fen Guo,Jiangxing Wang
DOI: https://doi.org/10.2991/ipemec-15.2015.10
2015-01-01
Abstract:Opinions' analysis on web news comments is a great significance for the construction of the government's democratic politics, public opinion monitoring and decision support, etc. In this paper we analyze user's opinion from emotional word, emotional word qualifier, and emotional tendency. We proposed an emotional word qualifier extract algorithm based on dependency relation tree. We improved the words polarity calculation algorithm based on HowNet. Finally, we implemented the system of opinion mining on web news comments, and got the overall sentiment analysis.
What problem does this paper attempt to address?