针对Apriori算法及其变种的不足,提出了一种基于支持度计数矩阵和事务数据库布尔矩阵的新算法,利用此算法挖掘石油期货价格历史数据中的频繁项集,根据给定的最小支持度和最小置信度从挖掘结果产生关联规则,对关联规则在石油期货价格预测中的应用进行了探索,并提出了进一步的研究方向. Aiming at the deficiencies of Apriori algorithm and its variants, a new algorithm based on support counting matrix and transaction database Boolean matrix is ​​proposed. By using this algorithm, the frequent itemsets in petroleum futures price history data are obtained. According to the given minimum support and The minimum confidence level generates association rules from the results of mining, explores the application of association rules in oil futures price forecast, and puts forward further research directions.