Intelligent decision support system for employee's performance prediction / Hamidah Jantan … [et al.]

The hidden and valuable knowledge can be discovered through data mining process. In data mining, classification is one of the major tasks to impart knowledge from huge amount of data. Knowledge discovered form data mining classification process can be embedded with Decision Support System (DSS) development which is known as Intelligent DSS (IDSS). IDSS uses Artificial Intelligent techniques to complement the work of human professionals. Nowadays, data mining techniques are widely used in various fields, but it has not attracted much attention people in Human Resource(HR) field. HR system is known as integrated and interrelated approaches to managing human resources and most of their activities involve a lot of unstructured processes such as staffing, training, motivation and maintenance. In addition, human decisions are subject to limitation where sometimes people forget the crucial details of a problem. Fair and consistent in evaluations are very important for HR professionals in any organizations. IDSS application using data mining can be used for evaluation, it will make many routine decisions in assessment easier and can be reallocated to less experienced evaluators. This system will encourages the use of explicit criteria for evaluating the employee performance, increases the assessment consistency, hence perceived fairness and provide help for junior evaluator to evaluate their staff consistently. This research focus on the execution of data mining approach for employee development regarding on their future performance. By using this approach, the performance patterns can be discovered from the existing database and it will be used for future performance prediction especially for their career development. In the experimental phase, we have used selected classification techniques to propose the appropriate technique for the dataset. An experiment is carried out to demonstrate the feasibility of the suggested classification techniques using employee's performance data. Thus, the experiment results, we suggest the potential classification techniques and the possible prediction model for employee's performance forecasting. Finally, the constructed model embedded in a system prototype for employee's performance prediction.