Design And Realization Of College Students' Comprehensive Quality Evaluation System Based On Data Mining

At present, quality education has been advocated by the whole society. With the development of the education and teaching reform, high education is endowed in the long-term and arduous task which upholding the core of quality education and cultivating person with high abilities and qualities. At present, high education has been come into a new period with the progress of the high education push forwarding quality education. However, quality evaluation especially comprehensive quality evaluation for college students seems to be weaker and the realism has been a bottle-neck affecting further strengthening and improving quality education.Comprehensive evaluation is objective, equitable and logical all-around evaluation to the object which has represented as the effect of quality education .It relates to the development of students themselves, the higher education teaching reform of college and the person of high quality which is acquired by the department. It is that weight is one of the important elements of the comprehensive evaluation, but actually the method of weight indexing is too single, subjective and unilateral, which could not objectively reflect its' importance. In order to avoid the shortage of previous evaluate system, the combinatorial weight method which has been come true through the existing system of subjective weight and objective weight choices and the favoritism of evaluators is adopted by the paper .The method has been absorbed in the advantages of objective weight and subjective weight and also avoided with the deficiency of two kinds of methods, and at last the weight reflects the importance of the indexes in the comprehensive evaluation more reasonable.The choice of comprehensive evaluation's method is very important, which is the most crucial step. Currently, the methods which can be better considered of the various qualitative and qualitative in the progress of comprehensive evaluation are described such as Artificial Neural Network and Fuzzy Comprehensive Evaluation. However, these methods of comprehensive evaluation couldn't get rid off the randomness, the subjective understanding of ambiguity and uncertainty and forth in practical application. Even if the same object often could be leaded to inconsistent results, as the same evaluation experts in different time and circumstances of evaluation. In order to avoid the deficiency of the past evaluation system methods and the system itself, various methods are combined in the system of college students'comprehensive quality evaluation. User could choose the appropriate methods based on the need of system. The methods of previous evaluation system is single, the evaluation results couldn't be comparable. The spearman method which is computed orderly and relatively by various projects, and selected the best project by the size of results is used in the system, moreover avoids the deficiency of it.Now, some comprehensive evaluation systems only pay attention to the last evaluation results, and the users use the results as the basis of estimating the evaluated object, they would rarely make use of the latent evaluation information, and then the system's degrees of integration and intellectualized are all lower. In order to not waste severity the resource of the valuable evaluation information, the technology of data mining is introduced into the paper. It is discovered that adopting the association rules in the evaluation data and attempted to find information in favor of decision-maker by the technology of data mining .It is a application of evaluation system which based on the college students'comprehensive quality in the paper and we can get the correlative information of whiz kid and dunce too. The decision-making department of college could find the excellent factors of whiz kid and the inferior factors of dunce from this, then suit the remedy to the case in the different students, which cultivate the higher diathesis students better and faster and more effectively.An improving association rules algorithm which based on vector plot is applied in the choosing the association rules algorithm of the paper. The algorithm only need to scan the database one time and would not generate a large number of candidate sets. It is firstly mapped data sets to a vector matrix and every vector represents a attribute of object. To judge the intersection of some frequent items, you can get the inner product by several representative vectors. If the final number of"1"is not less than accumulate of minimum support and the number of database affairs, the projects focus on the portfolio could merge to frequent sets. See section 5.4.3 specific algorithm. The efficient of vector algorithm much faster than Apriori algorithm which average times is between 10 and 20. When you mine a large number of undergraduate evaluation data, this algorithm may save algorithm a lot of time.The experiments show that those technologies of above paragraphs will be adopted into the comprehensive evaluation system of college students. The system is considered of qualitative analysis and quantitative analysis which has been provided with the characteristics, such as accord with the times and the needs of the job market. The system is deserved to generalize.