Mining And Upgrading Correlative Rules In Big Data Analysis

Under the background of big data decision,we develop the decision model in terms of the corrlatation of big data.We present concept of correlative decision system,and develop the algorithm of structure correlative decision count table in order to provide a data platform for the model behind the decision.Based on the basic algorithm of data mining,we improve the algorithms for mining correlative rule,and define the concept of correlative rules,then make construction to recommend a single commodity and multiple commodities at each time which suits the commodity recommendation system.According to the generated rules,it makes prediction and judgment,then provides the theory evidence for the specified industry of correlative decisions.In this paper,we have verified the correlative rules mining algorithm through experiments,and compared with the traditional association rule mining Apriori algorithm in terms of algorithm complexity,accuracy,practicability and generalization,and preliminarily proved the validity of the correlative rules.For the further spreading the application of correlative rules of algorithm,from the point of view of supermarket managers,we establish a correlation model between class and class.The correlative rule lifting algorithm is constructed to study the correlation between commodity classes.Finally,the conclusion is drawn and applied to supermarket management,which provides guidance for supermarket management decisions.The theory and method of this article can be further research with other fields,that will be getting more extensive promotion.