The Research On The Algorithm And Application Of Association Rules Based On Logistics Information

Association rule is one of the important fields of data mining. Using the strong correlation among commodities found by association rules algorithm can generate cross selling effect which will promotes enterprise profit, and speed up operation efficiency of inventory and other logistics services. This paper mainly pays much attention on the algorithm of association rules, puts forward improved algorithms to overcome the disadvantages of existing algorithm, and applies them to logistics practical to improve the efficiency of business decision. The main contents of this paper include:(1)After outlining data mining techniques, the shortcomings of association rules algorithm of the present stage are summarized:Firstly, classical association rules algorithms such as Apriori can only handle Boolean property, which are unable to handle quantifiable attributes contained in the actual data set. Secondly, as a measurement framework, support and confidence cannot extract strong association rules with low support value. And among the found rules, there may be false rules.(2)In order to overcome the disadvantage that traditional algorithm cannot handle the quantitative attributes, a quantitative association rules algorithm based on fast clustering is presented. The improved algorithm can deal with continuous attributes in the actual data set effectively by dispersing continuous attributes. It expands the application scene of traditional association rules algorithm. Actual data analysis shows that improved algorithm can effectively deal with numerical attributes and mine effective quantitative association rules.(3)In order to overcome the shortcoming that traditional framework cannot extract valid association rules, a new measurement framework to improve the evaluation effect is proposed. The framework is made up of Newrelevancy and NewI. Newrelevancy is used to find frequent itemsets and NewI is used to mine the strong association rules in these found frequent itemsets. Data analysis shows that compared to traditional measure framework, the new framework has a better evaluation effect.(4)Based on the above studies, in order to take advantage of the improved algorithm to improve the actual logistics operations, an improved ABC classification algorithm based on quantitative association rules is proposed. The improved algorithm sets H-confidence and relevance as the measurement framework to extract association rules and makes use of the Match degree to measure the cross-selling effect. Experimental results show there is a large difference on classification results between the improved algorithm and the traditional one. And the result of the improved algorithm has better practical application because it reflects the importance of commodity more accurately.