Twitter is one of the most popular social media where users can express their opinion and sentiment in a short message. Sentiment analysis is to identify a sentiment word from a given text and to estimate the overall sentiment orientation of the text....
Twitter is one of the most popular social media where users can express their opinion and sentiment in a short message. Sentiment analysis is to identify a sentiment word from a given text and to estimate the overall sentiment orientation of the text. It has attracted considerable attention in recent years because of its potential use in business and governments.
In this thesis, I introduce an approach to automatically analyze sentiment orientation of twitter messages that mention the current trends. An Internet newspaper is easily accessible and reliable information that deals with a current issue. From this Internet newspaper, I extract a set of topic words that are used to express the current issues or trends. By searching this set of topic words through twitter API, I can gather twitter messages containing the topic words. Sentiment analysis is performed on these twitter messages by using a sentiment lexicon. By doing this I can figure out the sentiment orientation of the twitter messages on the current trends. The empirical result presents that the proposed approach is useful for twitter sentiment analysis.