作者: Hao-Chiang Koong Lin , Min-Chai Hsieh , Wei-Jhe Wang
DOI: 10.1007/978-3-642-23456-9_83
关键词:
摘要: In this work, we collect the sentences posted in Plurk as our corpus. The emoticons are classified into four types based on Thayer's 2-D Model which is composed of valence (positive/negative emotions) and arousal (the strength emotions). system will preprocess sentence to eliminate useless information, then transform it be emotion lexicon. Besides, research analyzes three kinds semantic clues: negation, transition, coordinating conjunctions. final decided by SVM merging algorithm proposed work.