A Research Of Subject Evaluation Method Based On Data Mining

Data Mining techniques are useful in the situation of extracting information from a massof data and which is an important and effective means, through which one can obtain datafrom a large number of interesting and useful information.Currently, data mining technology has been applicated in business and other areas in aconsiderable size. However its application is not limited to this, it's known to us that the DataMining technology applicated in the field of education is relatively small, especially for theapplication of Subjects Quality Evaluation. Through the collection and mining of the data ofsubjects can draw a set of effective assessment tools, between disciplines so as to providehorizontal and vertical reference for the appraisal, and related disciplines to provide guidancefunctions.In order to realize a scientific and feasible subject evaluation system, we try to use thetechnology of Data Mining. The main technologies are the ID3algorithm in decision, treemethod and the Apriori algorithm in association rules. In this paper we give a briefintroduction of these algorithms and gave a method of subject evaluation based on them.In the first chapter of our paper, we introduce the research background of data mining,the international and domestic study status, proposed topics of background, the significance ofthe topics and the topics of practical value. And in chapter II, firstly we started with therelevant knowledge of Subject Evaluation, giving a brief introduction of the problem whichwe will talk about in this paper. In chapter III we elaborate the concept of Data Mining, theapplications of this technology, the task of data mining, process, classification of the basicclustering described and the description of knowledge as a basis for the preparation phase.The ID3algorithm of decision tree and the algorithm of Apriori in association rules as astarting point,a detailed discussion of the specific data mining algorithms to analyze how theattribute selection measure to take the concept of entropy in information on theory, and howto carry out tree pruning. Taking the shopping cart problem as an example, the relevant rulesof the association is explained and then introduced the concept of confidence and support todescribe the relevant Apriori algorithm. The chapter IV is focused on data mining in Subjectquality evaluation system. The application design and implementation process are describedaccording to the design of the program, the focus should be placed on the data pre-process work, the design of two data mining models the basic realization of a number of classificationwork. Finally in chapter V summarizes and discusses the evaluation in the subject field of datamining and knowledge of the feasibility of the method proposed in this paper shortcomingsand inadequacies and future work will be done to improve the local outlook, in order to live aguide to the future research.