Research Of Data Mining By Assocation Rules And Its Application To The Analysis Of Academic Achievements

With the development and wide use of computer and Internet technology, a large amount of data has been produced on people’s activities. The relationship between the data finding by analyzing and processing can help people make a decision or guide their social activities. Association rules come as one of the first problems of studying data mining, and the Apriori algorithm is the most classical algorithm of association rules, which can be widely used in a variety of scenarios. But as the algorithm needs to access the database frequently and the support-confidence degree is redundant, the time efficiency of the algorithm and the accuracy of the information would be low. This paper mainly studies and improves the Apriori algorithm by reducing the algorithm’s time and improving the accuracy of association rules on the basis of the Apriori algorithm.In terms of time efficiency, the improved algorithm mainly contains two aspects, one is to reduce the times of accessing database; the other one is to reduce the times of scanning transaction. In order to reduce the frequency of scanning the transaction in database, the improved algorithm maintains a Map table instead of scanning the transaction in database every time when calculating the support of the large item-set, the Map table is used to record the number of transactions ID where the frequent item-sets are located, the support count of frequent item-set is the intersection of two subsets of the set. In order to reduce the frequency of scanning the database, the transaction database can be split into several disjoint parts by scanning each partition of the transaction, we can obtain local frequent item-sets when scanning each part, the candidate frequent item-sets come from all the local frequent item-sets, we can get frequent item-sets by re-scanning database.In order to improve the accuracy of the information, improved algorithm adds the importance and validity as a measure of the standard in the process of generating association rules, no longer relying on the single standard named the Support and Confidence. According to the calculation of the confidence, we come to the conclusion:through some form of deformation of an association rule, association rules derived must be strong association rules, which is not what users interest and needs to be deleted. Therefore, we should improve the algorithm in depth to a higher level.By comparing the improved Apriori algorithm with the traditional Apriori algorithm and AprioriTid algorithm, the superiority is proved both in time efficiency and information accuracy. Finally, the improved one is applied in the analysis system of students’ achievement by dealing with rules between student achievements, and analyzing and interpreting the results of association rules. The rules can be used to warn students and help them bear learning tasks in mind. Meanwhile, teaching staff can use the rules to make reasonable training plans, and thus improve the quality of teaching. Meanwhile, association rule has a very good practical significance in its application in achievement analysis.