Abstract: Aiming at the the inefficiency of traditional frequent itemsets mining algorithm in view of the big data environment, a solution to this problem parallel weighted mining of frequent itemsets using PWFIUT(Parallel Weighted Frequent Itemset Ultrametric Tree) algorithm is implemented on MapReduce framework. Support is counted by mapping the items from the candidate list into the buckets which is divided according to support known as Hash table structure, also to avoid building conditional patterns and to achieve compressed storage. Finally, the Algorithm is verified and analyzed on Hadoop platform. According to the compared experiment results, it shows that the proposed algorithm has high efficiency and good scalability.