Development of Requirement Elicitation Model for Prediction of Student Achievement in University

Human needs of information increase periodically. These improvements lead to knowledge discovery which can be used to prediction. One of the organization that needs prediction to support their business processes is college or university. Universities need to predict their student achievement as support of their business objective. In fact, the software owned by universities have not been able to help their stakeholder in prediction, especially in predicting student achievement that based on the accuracy of graduation time. In addition, university's data are very huge and increase periodically which raised the need to implement specific techniques in developing software. This techniques called data mining classification. The software that will be built requires a software development process that is quite different from the other types of software development, especially in terms of requirement elicitation. This causes the existing requirement elicitation models need to be customized to the characteristics of business process in universities and implementation of data mining technique. To solve these problems, in this research was developed requirement elicitation model that specific to the needs of prediction in universities. Requirement elicitation model is then used to form a general solution model of prediction using data mining classification techniques in universities. Models that have been developed become the basis for the software development that can be used to predict student achievement in universities. Based on this research was obtained new requirement elicitation model that have been customized to characteristic owned by universities and the need to implement data mining classification technique. After develop new model, a software was developed for the prediction of student achievement as the implementation of the model.