Reader Emotion Prediction Using Concept and Concept Sequence Features in News He

This paper presents a method to predicate news reader emotions. News headlines supply core information of articles, thus they can serve as key information for reader emotion predication. However, headlines are always short which leads to obvious data sparseness if only lexical forms are used. To address this problem, words in their lexical forms in a headline are transferred to their concepts and concept sequence features of words in headlines based on a semantic knowledge base, namely HowNet fo