In order to explore the eutrophication evolution mechanism and identify the key factors of water environment that affect the eutrophication and cyanobacteria growth of Taihu Lake, the data preparation and data cleaning for the data of the multi-source monitoring sequence of Taihu Lake from 2006 to 2018 were carried out. The K-means clustering method was used to obtain the discrete Boolean association rule for the mining of candidate data sets, and a mining model of association rule for key factors of Taihu Lake was constructed based on the Apriori algorithm, from which the key factors of water environment that affect the eutrophication of Taihu Lake were identified. The results showed that the mass concentration of chlorophyll a, which characterizes the degree of eutrophication in Taihu Lake, has different degrees of correlation with total phosphorus, ammonia nitrogen, pH and permanganate index. Among them, the mass concentration of chlorophyll a in the range of 0-18. 36 mg/m³ has the strongest correlation with the total phosphorus in the range of 0-0. 045 mg/L. From the perspective of water environment management, if the total phosphorus concentration in Taihu Lake is controlled below 0. 045 mg/L, the probability that the mass concentration of chlorophyll a in the whole lake is below 18. 36 mg/m³ would be the highest, which can effectively control the number of cyanobacteria in the overall state of less and further avoid the large-scale outbreak of cyanobacteria bloom in Taihu Lake.