Abstract
There is much news online every day and people would like to read online news and give their comments on some news, which reflect their views and opinions about the people and matters involved in the event reported in the news. Opinions mining on these comments is of great significance for the construction of the government's democratic politics, public opinion monitoring and decision support, etc. In this paper we analyze user opinion from four aspects: characteristic, emotional word, emotional word qualifier, emotional tendency. We propose aSRL-DP-NEalgorithm to extract the characteristic-emotional word pairs. Finally, we implement the system of opinion mining on web news comments, and got the overall polarity analysis.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have