An Approach of Building Microblog Smiley Emotion Lexicon and Its Application for Sentiment Analysis
Lin Wang · Computer and Digital Engineering · 2012
As a new network language,smiley has been widely used in microblog and can represent the user's mood and thought to a certain extent.Moreover,smiley will affect the result of the microblog sentiment analysis.In this paper,an idea building sentiment lexicon for the smileys is proposed based on statistics data of microblog.By analyzing sentiment orientation of texts co-occurring with smileys in microblog,the sentiment orientation of every smileys is assigned.Based on the approach,a smiley lexicon is built for sentiment analysis,so as to support the sentiment analysis of microblog and other user generation information with smileys on Web.Moreover,the smiley lexicon is fed back to their microblog texts for re-measuring orientation value of words in the texts and improving existing sentiment lexicon,so as to acquire more accurate sentiment analysis results.Experiments show the effectiveness of this method and analyze the related threshold setting on the influence of the results.