Data Mining-Basedunemploymed Reemployment Prediction Model Research

The employment and reemployment are keystones in China in recent years. All sectors of society attache great importance to employment work and explore a lot of useful experience. But with urbanization, industrialization and economic restructuring process, labor relations will be more complicated.This paper is supported by Science and Technology Department of Zhejiang Province. It's the core research of "Grid-based Distributed Labor Market Decision Support System", and mainly on the labor market-oriented decision support functions layer. This paper begins with the research and development of theory about unemployment at home and abroad, China's labor market characteristics, status, problems and solutions. By summing up and analysing, the prominent issues of reemployment of unemployee such as supply and demand of the unemployed labor, the unemployment rate, the reemployment guidance are conducted in-depth. This paper takes data mining and other innovative techniques into applications. It mainly constructs a trend-based time series prediction model for forecasting supply and demand of labor, a competitive neural network-based unemployment early-warning model and a cluster-based reemployment difficulties score rating model.The data of the human resources system of Hangzhou Labor and Social Security Bureau and Statistic Annualis of Hangzhou in the 2006 are used as the input of these above models. The open-source Data Mining software Weka is used as the experiment tool. These prediction models and early-warning models provide a quantitative analysis and decision support methods for unemployees to find new jobs.