期刊名称：International Journal of Computer Science & Technology
印刷版ISSN：2229-4333
电子版ISSN：0976-8491
出版年度：2011
卷号：2
期号：3(Version 1)
出版社：Ayushmaan Technologies
摘要：Human Emotion Recognition (HER) in text documents is essentially a content – based classification problem involving concepts from the domains of Natural Language Processing as well as Machine Learning. Apart from HER and Human Computer Interaction (HCI), classification of text documents on the basis of their content also finds use in fields like context – based Information Retrieval (IR) and Text To Speech (TTS) conversion. A number of techniques from Machine Learning have been employed for achieving emotion – based text classification, like Decision trees, Na.ve Bayesian classifiers, Support Vector Machines and Neural Networks. However, an important part of the classification algorithm is the appropriate representation of the text documents for input to these classifiers.s
