Abstract: Online Banking system refers to manage the whole banking system through web. In banking system, it is necessary to retrieve important information of customers from their historical data to give them some opportunities, such as discount on loan interest. To do this it is very necessary to classify them. We have classified the customers based on classical logic and fuzzy logic. We have seen fuzzy classification provides better performance than classical classification. In the last decades, information systems have revolutionized the way information can be stored and processed. As a result, the information volume has significantly increased. It becomes difficult to analyse the large amounts of available data and to generate appropriate management decisions. In practice, Information systems mostly use relational databases in order to store these data collections. This paper makes a comparison between traditional or classical classification and fuzzy classification. Experimental results demonstrate that the proposed intelligent fuzzy query is more effective than the conventional query and it provides the user the flexibility to query the database using natural language. Traditional online banking system is not intelligent, because we can’t retrieve the behaviour of customers through the system. Online Banking system refers to manage the whole banking system through web. It leads to the virtual office. Data mining is the most important research work in real world application. We can retrieve knowledge of employees and customers of any organization from their historical information through data mining. Online banking system with data mining is intelligent. In my research work, I include the most important data mining task classification and I have implemented the classical classification and fuzzy classification on online banking system. And finally, I have showed the comparisons between classical classification and fuzzy classification. Consequently, I have seen the fuzzy classification is better than classical classification due to the over and under estimation value closer to the boundary of the intervals

Keywords: Include at least 4 keywords or phrases