Analysis And Evaluation Of Preference Selection For National College Entrance Examination Based On OLAP And Data Mining

On January 11,2008, the Ministry of Education asked to improve the Preference Selection for National College Entrance Examination(PSNCEE), encourage all the provinces using parallel Preference Selection from "The notification of National College Enrollment in 2008". The reform of parallel Preference Selection has spread out, changing from "Preference Priority" to "score priority and follow preference".However, the social community, especially all the examination institutes and students concern the rationality of the new PSNCEE and whether it can satisfy all the students, especial excellent students'preference. It is an urgent issue for evaluating the PSNCEE in a scientific, systemic and quantitative way, in the mean time, providing a better PSNCEE.According to the data from National College Entrance Examination of one province for many years, this thesis arranges mess data using Data Warehouse, and evaluates parallel Preference Selection by means of preference coverage rate, preference matching, preference fulfill rate and the lowest admission line for latest years through OLAP. Besides, this thesis also evaluates it with Data Mining through three aspects including the top three preference enrolling situations,2005 split points etc.The result indicates that the "1+2+3" parallel Preference Selection in 2005 protected excellent students in a certain extent, increased the satisfaction of PSNCEE.Also the relative research may provide theoretical principle for "One Record for Many Preference Selection", and the decision support for recruit institutions of every province.