In this paper, we propose a new classification algorithm called Intelligent Agent and Neuro-Fuzzy Rule based Group Support Vector Machines (IGSVM) to perform major classification of sentiments and to form groups based on the sentiments of people with respect to change in time and place. Finally, the groups are used to form discussion forums on various topics including business, e-learning, tour and sports. The main advantage of the proposed work is to identify the user interest based on the sentiments identified from tweets and to form similar interest users groups for discussion on specific topics. From the experiments conducted in this work, it is proved that the user groups formed by sentiment analysis provided more than 94% accuracy in identifying members for forming interest groups on twitter and hence is more accurate than the existing systems.