Frequent and high utility itemset mining is an important task in data mining, which aims to mine a set of frequent and high utility itemsets, and the mined itemsets are measured by two metrics, support and utility, to determine whether they are frequent and high utility, respectively. Among a series of methods used to solve such problems, evolutionary multi-objective methods have achieved good results, providing a set of high-quality solutions to meet the needs of different users, as well as avoiding the problem of difficulty in determining the threshold values of support and utility in traditional algorithms. The existing multi-objective algorithms are encoded with 0-1 and the dimensionality of the decision space is proportional to the number of items in the dataset. Therefore, the curse of dimensionality problem can occur in high-dimensional datasets. In order to solve this problem, this paper designs an itemset reduction strategy to reduce the search space by reducing the unimportant items to solve the dimensional catastrophe problem. According to this strategy, the article goes on to propose a high-dimension frequent and high utility multi-objective optimization algorithm IR-MOEA for itemset mining based on itemset reduction, where a learning-based population restoration strategy is proposed to adjust the evolutionary direction for over-reduced or under-reduced individuals. In addition, an initialization strategy is proposed to generate sparse solutions that facilitate evolution at the early stage of evolution. Finally, experimental results on multiple real and artifical big datasets show that this algorithm outperforms the existing state-of-the-art multi-objective optimization algorithms for mining frequent and high utility itemsets, especially on high-dimensional datasets.