Research And Implementation Of Human Resource Management System Based On Data Mining And SOA

With the development of economic globalization, In China, large and medium enterprises play an increasingly important role in the international competition. A higher request was promoted to China's information management level. As the most important strategic resources of a group, enterprise human resources management also needs to be in line with the world-class management level. To guarantee that human resources with the technology environment which is altering from day to day develop simultaneously. Optimizing the process of human resources management, implementing the automation of human resources management roundly and the transformation from traditional human resources structure to human resources information, are hot topics in information technology field research, designs and implementation at present.Service-Oriented Architecture (SOA) is a kind of architecture, which is centering on serving. It is an abstract concept and the guidelines of software architecture, and is a bridge between the information technology and the concrete business. Through it's a series of design principle, design pattern and implementation technology, the concrete business information service can be achieved on. Simultaneously, the coupling of systems is reduced in most extent and the reuse of software is enhanced.Human resources management is regarded highly by each field gradually as one kind of management pattern which is an important component of advanced business administration. Human resources management is facing the mass data, which needs one kind of technology to discover the valuable knowledge urgently, to formulate human resources developmental strategies and to provide support for decision-making.This dissertation is mainly discussed the application of service-oriented architecture and data mining technology in the human resources system. Through the massive exploration and practice, lightweight layered architecture based on SOA is proposed, as well as in view of the ID3 algorithm when test attribute are selected, the impact of the correlativity over attribute choosing a testing is not take into account. So, a pre-pruning ID3 algorithm based on the mutual information between attributes which is to build better decision tree is proposed. In this dissertation, the requirement analysis and modeling to human resources management system are carried on. It takes the performance evaluation module in human resources management as the example, the system provides the auxiliary decision-making for the policy-making support by data mining technology for the analysis of existing data, processing. Thus it makes human resources management quality enhanced remarkably and is more flexible adapt to the modern human resources management needs.