Abstract: In a hybrid cloud environment, data flow has disturbance, the anti jamming problem of adaptive compression mining algorithm for scattered point cloud is not good, a hybrid cloud environment for extracting data information flow is proposed based on phase space reconstruction and association rule mining algorithm under the characteristics of data stream. First, to construct a hybrid cloud environment information data flow time series analysis model for data structure analysis, and then the hybrid cloud environment data flow information flow for high dimensional phase space reconstruction of association rules in the feature extraction phase space reconstruction, with the extracted features as pheromone positioning guide data mining, data mining algorithm is improved. Finally through the simulation experiment of performance test, the results show that, by using the algorithm in hybrid cloud data mining features extraction has better convergent, higher accurate probability of data mining, anti-interference performance is better than traditional methods.