Neuro-Fuzzy Based Intelligent Agent for Event Based Emotion Recognition

Emotion recognition from text is a necessary step to develop affective conversational interfaces. Intellect cannot work at its best without emotional intelligence. This paper deals with the development of an Intelligent Agent that exhibits the concept of emotional intelligence. The approach chosen for the implementation is Soft Computing and the architecture used is a Neuro-Fuzzy system the input to the system is a real life event which is tokenized and the tokens are compared to a corpus of emotional keywords. Based on the matching, the emotional values of the keywords are computed and are processed by the neuro-fuzzy controller to generate the corresponding emotion underlying the event. The network is trained using the backpropagation algorithm. The entire system is implemented in C++ framework.