Application Research of Parallel Frequent Itemset Mining Algorithm in Book Recommendation

Based on the traditional FP-Growth algorithm, this paper designs an improved Spark FP-Growth algorithm and deploys the improved serial algorithm to the Spark cluster to realize the parallelization of the algorithm. In this paper, the IPFP algorithm is used to mine frequent itemset from the library borrowing database and generate association rules. Then the generated association rules are used for the library's book selection recommendation service, which realizes the transition from the traditional book-oriented to the user-oriented, thus improving Work efficiency enhances the initiative and individuality of book selection.