College Entrance Examination Network Reported System Performance Optimization And Volunteer Forecast Analysis

Along with the rapid development of computer technology, the issue of how to gather the college application of college entrance candidates timely and accurately and provide a voluntary choice of college application and professional guidance is that the candidates, parents, schools and enrollment management departments are concerning about.Select a subject of analysis and optimization of the performance of college application system and the forecast and analysis of the college application for this paper. For the fact that there is huge visitor volume and high concurrency during the process of filling in college application form via internet due to the concentration of filling-in time, through the use of Web clusters, database clusters, procedures, and database optimization to improve system performance. Refer to applying to college for the prediction, the use of statistics learning style, through analysis of historical enrollment data, build forecasting model Entrance Examination for candidates with voluntary reporting decision-making. Specific research papers are as follows:For the analysis and optimization of the performance of college application, the fact that there is huge visitor volume and high concurrency during the process of filling in college application form via internet due to the concentration of filling-in time. The paper puts forward a method to optimize the performance of college application system. It intends to meet the requirement of high concurrency of system from hardware platform, Web service as well as database service via constructing Web cluster service by Web integrated approach and constructing database cluster service by Oracle RAC service. It intends to promote the execution efficiency of program, promote the supporting capability of single Web service as well as database service, and finally promote the overall efficiency of system via optimizing SQL program, table index, associated multi-tables, as well as the analysis of program binding variable database.For the prediction analysis of the College Entrance Examination, the proposed method using support vector machine learning methods to extract results of the candidates, candidates are ranked categories of candidates, the number of college enrollment plan the institutions of the quality characteristics of the 21 features, using a province of 18 million admissions last two years of historical data, training of different batches of College Entrance Examination constructed prediction model, and conducted experiments comparing the classification of different characteristics, the results show that the combination of candidates information on the characteristics of information and institutions build classifier achieved good results.Online reporting service for the College Entrance Examination, the use of J2EE architecture, design and implementation of the College Entrance Examination network reporting system, and optimized performance, and Yunnan Province, College Entrance Examination in 2010, reported the application of the network, active link peak to 27 million, the entire system is running stable, ensuring the effective voluntary college entrance candidates, timely and accurate collection. Prediction in the College Entrance Examination, the use of J2EE architecture, design and implementation of the College Entrance Examination predict the prototype system, individual schools for candidates to declare for predictive analysis and recommend candidates to fill the information of the College Entrance Examination School Prediction of the voluntary and recommendation work.