Automatic collection polarity adjectives in sentiment analysis
Accompanied with the explosion of Web 2.0 platforms such as blogs, discussion forums, and other types of social media, sentiment analysis has become an active research topic in recent years. In order to expand high quality polarity adjectives as many as possible by using WordNet, we propose three types of query string in this paper. Among them, the third type of query string is the most important. Besides 'synonyms' and 'antonyms' relations, we also use 'similar to' and 'also see' relations of WordNet. In fact, the last two relations are more important than the first two ones. Once taking each synset as a special kind of document, we can rank those expanded adjectives by using the X 2 test. Extensive experiments demonstrate that our POAE algorithm is very effective for expanding polarity adjective seed list in sentiment analysis.