Abstract:

The aim of the paper is to introduce the design of an intelligent social agent which interacts naturally with the user by analyzing and responding to both contextual and affective cues obtained through text or speech and bio signals.  Generally, the user of the system will be wearing a pervasive device which sends readable bio signal to the system.  The bio-signals received from the user will be analyzed by an emotion component which will use fuzzy logic to classify the ‘real-time’ emotional states of the users.  The emotion states will then be used in tandem with the direct cognitive input from the users to analyze the ‘precise’ intended meaning of the users’ responses.  The results of the analysis will then be processed for an appropriate response from the behaviourial system based on the selected personality traits which could be customised.