Datamining In Teaching Appraisal Applied Research

Datamining is the process of abstraction unaware, potential and useful information and knowledge from plentiful, incomplete, noisy, fuzzy and stochastic data, which is deemed to one of a foreland of datamining system and a promising cross-subject. There is a good many question at the practice teaching appraisal of them, the tradition means of teaching appraisal is disability in the face of a great deal teaching data of accumulation for past year. Datamining solves difficult problem. The Cluster method is one of more important role in datamining. This dissertation systematically and deeply studies and analyses the datamining technique, especially the one for association rules, furthermore applies it to teaching appraisal. The main contents are listed as follows.At first, the appearance of the datamining technique is reviewed in brief. Based on the basic concepts of datamining, this dissertation not only classifies and summarizes the findable patterns of datamining in detail, but also studies architecture structure and running process of datamining. In succession, the dissertation summarizes and studies the current status of the datamining technique in our native country and overseas. All of the above become the basis for this dissertationThen, the dissertation discusses The Cluster methods for association rules. k mean partition algorithm and layer Clustering algorithm are the one of emphases contents in datamining. The two algorithms are appliance in abroad. However, with deeply application, The shortage emerges in the two algorithms .For example, there is calculation of the distance of the single link when subarea and combination for the Layer Clustering Algorithm. It uses plentiful time. Its time limit is o(n2). Besides, when a combination is finished, it is not retracted. However, no designation subarea count in advance is the most excellence for k mean partition algorithm. The count k of enactment subarea is designated in advance,which is impertinency for the user .It will be constringency to be a least rule of the part when the begin subarea is improper ,for k mean artition lgorithmt bring out to find the excellent answer. Aiming at the instance,the paper points out an improved new Algorithm (NP Algorithm).It applies the datamining to teaching appraisal for the adult ducation of WEIFANG college base on dataing tool of MS Analysis Services.it obtains the student characteristic thorgh decision tree analysis and Clustering analysis. For instance, the synthesis representation of the schoolgirl's study grade, which age is less than or up to thirty, is lean to middling. The synthesis representation of the schoolboy's study grade, which age is more than thirty, is lean to excellence etc.Those characteristic will be provided with learning counselor. They will find the problem and the disciplinarian. They will analyse distinctness characteristic. They will bring forward the guidance deliverance. (Data is from education professional of the adult education of the WEIFANG College.)The recent data of Educational Technology Major WEIFANG College are analyzed with SPSS technology, some information and knowledge are mined, and some education phenomena are discovered, for instance, the schoolgirl and schoolboy exist in salience difference at the graduate paper.These instance happens at the grade examination, which grade of schoolgirl is high in evidence those of the schoolboy, which support to educational decision-making. Based on their research, I put forward a model of teaching appraisal which is adapted to WEIFANG College.