The Application And Research Of The Paper Analysis Algorithm

Examination is one of the most important parts in teaching, which is the most effective method to feedback teaching information, and to test teacher's teaching level and students'learning circumstance. As an important carrier of examination, the papers have drawn more and more attention to analysis and to assess their quality.It must be a detail analysis of the papers, to evaluate which whether achieve the purpose or not. At present, paper analysis usually gets information on the surface by counting the scores simply, while overlooks a lot of valuable information which is hidden in the papers deeply. Moreover, paper analysis system available is mostly single in function, and paper analysis which is becoming a form can't meet the requirements of modern teaching management. In order to solve the problem of deep mining valuable information, this paper made the following aspects of work:First, according to educational measurement theory, combining with the paper quality analysis and "four times" theoretical model of assessment, establish a standardized evaluation papers system, identify the indicators of paper analysis, and provid a basis for an accurate and scientific assessment of papers.Second, study the basic theory of association rule mining, including basic concepts of association rule mining, classification, data mining process, commonly used algorithms and so on, then make a comprehensive analysis of classic Apriori algorithm and AprioriTid algorithm in association rules, and point out the inadequate and propose an improved AprioriTid algorithm.Finally, design and realize the paper analysis system, including the performance of conventional and network inputting, the analysis of paper and question, performance analysis, association analysis and other functions. Make effective evaluation of the paper, create paper quality analysis and assessment report.The improved AprioriTid algorithm is applied to paper analysis system, the experimental results showed that:through analysis of student papers, the method find correlation between the course, and based on questions of scoring rate, identify the relationship between questions, get a meaningful rules.