Abstract: Due to the data feature correlation in the initial item set, the data mining result of the association rule Apriori algorithm exists error. In order to solve this problem, an improved association rule data mining algorithm was presented based on rough set theory (RST). Then, the algorithm was applied to software engineering risk factors and risk mitigation factor management analysis, and a new software engineering adaptability structure was proposed. The simulation results showed that this improved algorithm increased the efficiency of mining data.