i-SME: Loan Decision Support System Using Neuro-CBR Approach

The granting of loan by a financial institution is one of the important decision that require insubstantial care. Currently, SME Banks in Malaysia are using conventional approach to process loan applications. Due to this conventional approach, analyzing information related to entrepreneurs manually is very time consuming. Based on SMEs expert criteria in Malaysia that has been collected from corporate sector and financial institutions such as SME Corp. Malaysia and SME Bank, the design and development of the prototype for an intelligent decision support system has been implemented. Therefore, this study extends the manual concept of SME loan application processing to a digital, automated and intelligent processing that learns and supports the user in decision making. It explores the use of hybrid technology such as Neural Networks and Case Based Reasoning. The system known as i-SME is able to classify SME market segment into three distinctive groups, there are MICRO, MEDIUM and SMALL with accuracy of 98.97 percent. In addition i-SME recommends to the management whether an application should be accepted or rejected. The evaluation based on Perceived Usefulness and Perceived Ease of Use reveals that i-SME is useful and easy to use. Furthermore, it is possible to transform the patterns generated from i-SME into actionable plans that are likely to assist SME Bank to be more effective and competitive.