Decision Tree Based On Data Mining Technology To Performance Evaluation System

The most important thing in21st century is talents. For an enterprise, undoubtedly the core competitiveness is to have a strong team full of talents, and to attract and retain these talents, an enterprise needs to have effective performance evaluation and incentive system. Talking about the management, only with effective performance appraisal under the local condition and recognition to the staffs work, can the enterprise let their staff have a sense of identity and belonging, and further more, the enterprise can seize the initiative in the market competition.Currently, many IT companies only have their eyes locked on the business expansion and the company’s market share,but not on the effective employee performance evaluation, either way, they just accumulated large amounts of data for a simple backup, and query statistical operations, instead of doing an appropriate analysis of the causes of performance evaluation results, so they can not improve for the evaluation of the problems.Data mining can be conducted in in-depth information on data mining analysis, apply it in performance evaluation can be beneficial to explore the potential inner link between the evaluation results and various factors. Data mining can excavate useful information from historical data, and can show them up in the form of classification rules, thus to make better use of data accumulation, which will help us to improve the quality of performance assessment and improvement.This paper presents a new research program on the application of the decision tree of performance appraisal system, and dig out the valuable hidden information in the back of the staff performance appraisal. It can identify the core attribute of staff performance appraisal through a large number of real data mining process, for whose purpose is to effectively guide the work of the staff to enhance operational efficiency, and improve staff’s work ability.Also this paper describes the details of data mining and decision tree classification concepts and methods. Through the evaluating-and-comparing of common classification algorithm, we choose to use the decision tree classification algorithm in the research; by comparing the advantages and disadvantages of several decision tree algorithms, we choose to use ID3algorithm for mining research, according to the characteristics of each algorithm, and the actual situation of performance evaluation data. In addition, this paper improves the traditional ID3algorithm for multi-value bias problem and also proves the effectiveness of the new algorithm. Combined with employee performance evaluation data provided by company X,we use the improved algorithm to do an classification analysis of the evaluation of data, from which we present the decision tree analysis model of performance evaluation based on applied research.