Research on Mining Sequential Association Rules Based on Conditional Confidence

Different process sequences are involved in the assembly process of aerospace complex components, among which there are many common and typical process sequences. But at present, the reusability of a large number of process files is low, and the reuse of existing process files cannot be realized. Therefore, mining out the typical process sequence can be used to guide technologists to write new process, realize the efficient reuse of historical process scheme, effectively improve the efficiency of equipment process preparation, and shorten the process design cycle. Based on the traditional Apriori algorithm and FP-Growth algorithm, this paper studies a sequence based association rule mining algorithm—Sequential Frequent Pattern Growth (SFP-Growth Algorithm). The algorithm retains the advantages of FP-Growth algorithm in constructing FP-Tree, and learns from the characteristics of Apriori algorithm in generating frequent items. The algorithm has a slight improvement in the efficiency of the algorithm. Compared with FP-Growth algorithm, it also has an improvement in the acquisition of frequent itemsets. And it can excavate the potential sequence relationship between processes, which can be better applied to the process design system and provide great convenience for the process personnel.