Title :
Microblog Orientation Analysis Based on Connnectives
Author :
Jun Zhao;Hong Wang
Author_Institution :
Sch. of Inf. Sci. &
Abstract :
At present, a larger number of researchers analyzed Micro-blog orientation and they concentrated their energy on emotional words, adverb and negative words without considering the impact of other emotional factors, such as the coordinative relation, progressive relation and adversative relation. So a method studying the sentiment polarity of associated words was proposed. The structure characteristics of associated words were analyzed sufficiently, and a turning words dictionary, a Chinese emotion dictionary and a negative words dictionary were founded based on the existing dictionary. This paper considers the combination laws of negative words, adversative words and conjunctions into Micro-blog. In the experiments, we verify this method can analyze the sentiment polarity of transition sentences efficiently and also has higher classification accuracy.
Keywords :
"Dictionaries","Support vector machines","Blogs","Turning","Feature extraction","Mutual information","Software"
Conference_Titel :
Broadband and Wireless Computing, Communication and Applications (BWCCA), 2015 10th International Conference on
DOI :
10.1109/BWCCA.2015.40