CAJ | 학술논문

对海量教育信息化数据关联模式的挖掘与分析，有利于发现教育信息化数据中事先不知道的、但又是潜在有用的教育模式和知识，为教育管理、教育决策和教学改革提供科学的依据。研究面向教育信息化数据的完全加权负关联模式挖掘技术，提出一种新的教育信息化数据完全加权正负关联模式挖掘算法。该算法考虑了项目在事务数据库中的权值，从教育数据中挖掘完全加权频繁项集和负项集，在支持度-置信度-相关性架构下从频繁项集和负项集挖掘有趣的教育数据完全加权正负关联模式。通过模式分析，发现教育领域潜在的教育、教学规律和教育发展趋势，提高教育信息化数据分析和处理水平。以真实的教育信息化数据作为实验数据测试集，实验结果表明，与现有的正负模式挖掘算法比较，所提出的算法更有效、更合理，其挖掘出的正负关联模式更加接近现实。
대해량교육신식화수거관련모식적알굴여분석，유리우발현교육신식화수거중사선불지도적、단우시잠재유용적교육모식화지식，위교육관리、교육결책화교학개혁제공과학적의거。연구면향교육신식화수거적완전가권부관련모식알굴기술，제출일충신적교육신식화수거완전가권정부관련모식알굴산법。해산법고필료항목재사무수거고중적권치，종교육수거중알굴완전가권빈번항집화부항집，재지지도-치신도-상관성가구하종빈번항집화부항집알굴유취적교육수거완전가권정부관련모식。통과모식분석，발현교육영역잠재적교육、교학규률화교육발전추세，제고교육신식화수거분석화처리수평。이진실적교육신식화수거작위실험수거측시집，실험결과표명，여현유적정부모식알굴산법비교，소제출적산법경유효、경합리，기알굴출적정부관련모식경가접근현실。
Mining and analysing the association patterns of vast amounts of educational informatisation data is helpful in discovering new and potentially useful educational patterns and knowledge from the data,and it may also provide a scientific basis for management,decision-making and teaching reform in education.In this paper,we probe the technique of mining all-weighted negative association pattern for educational informatisation data,and put forward a new algorithm of mining all-weighted positive and negative association patterns from educational informatisation data.The algorithm takes into account the weight value of the items in the transaction database and mines all-weighted frequent itemsets and negative itemsets from educational data,and then mines interesting all-weighted positive and negative association patterns from these itemsets in the support-confidence-correlation framework.Through analysing these patterns,one can discover the potential educational and teaching rules in education sector as well as the educational development trend,thus can improve data analysis and processing levels on educational informatisation data.It is also demonstrated from the results of experiment on real educational informatisation data that compared with existing mining algorithm on positive and negative patterns,the algorithm proposed in this paper is more effective and reasonable,and the positive and negative association patterns mined by the method get closer to the reality.