Research On Emotional Information Recognition Of Chinese Text Based On Hownet

With the development of Internet, textual information becomes more and more, and becomes the richest interactive resource. In the face of abundant affective information resources day by day on the web, how one can obtain and utilize the information fast and effectively is becoming a question that people pay close attention to. Text is a particularly important modality for sensing affect. The bulk of computer user interfaces today are textually based. Textual emotion recognition means recognizing the potential affective information from the text. It has become the key part of human-machine interaction.The main contributions of this paper are:This paper presents a novel way for building an emotional thesaurus based on HowNet, which is an on-line common sense knowledge base unveiling inter-conceptual relations and inter-attribute relations of concepts as connoting in lexicons of the Chinese and their English equivalents, and using the emotional thesaurus makes some research on the emotional information recognition of Chinese text.In the process of building the emotional thesaurus, first, according to the current development of the emotion classification and the emotional information contained in HowNet, the emotion classification system is confirmed. Then according to the inside structure of HowNet and inter-conceptual relations and inter-sememe relations of HowNet, with the help of manual tagging, the emotional thesaurus is built. After that, with NLP techniques, using the emotional thesaurus, the emotional information recognition system is implemented, and some experiments and result analysis are made at last.