Logistics Customized Service Push System Based On Association Rules Analysis

With the advent of the era of big data,the process of economic globalization is accelerating,the competition among enterprises is increasing,the competition of logistics services is also intensifying,and the logistics service delivery system is also applied.In the context of service globalization,market competition between logistics service delivery systems is also increasing.How to provide more accurate and meaningful service push is extremely important for the logistics industry.From the customer's personal basic information,historical records,etc.,to construct a customer portrait of the customer's demand behavior,based on this,a series of service pushes,which can effectively solve the mechanical and blind existence of the previous logistics customized service recommendation process.Sexual issues that increase customer loyalty and satisfaction.In this paper,the association rule analysis technology is used to process the relevant data of logistics customers,and the relationship between customers and commodities is obtained,and on this basis,logistics customers are provided with better recommendation services.The recommendation technology based on association rule analysis and collaborative filtering algorithm is two methods applied in personalized recommendation.In the traditional association rule analysis algorithm,it is difficult to find rules,occupy unnecessary system resources,and take time.Longer and other shortcomings,in the collaborative filtering recommendation algorithm,there are also problems such as data sparsity,cold start and so on.Therefore,this paper proposes a recommendation algorithm that combines an improved algorithm based on association rule analysis with a collaborative filtering algorithm to compensate for the shortcomings of the two basic algorithms.In this paper,we add the idea of weight and high expectation probability to the association rule algorithm to improve the efficiency and accuracy of rule extraction,while reducing system overhead,but in the face of new project problems,this method can not Effectively solve the problem,so we propose to combine the association rule improvement algorithm with the collaborative filtering algorithm.By weighting the results of the two algorithms,the Top-N method is used to achieve the final recommendation result.This not only enables collaborative filtering to effectively improve the new project problems existing in the association rule recommendation process,but also improves the data sparsity of collaborative filtering in mining potential rules.In summary,this paper improves the association rule analysis algorithm and combines it with the collaborative filtering algorithm to make the recommendation result more accurate and efficient,and at the same time diverse.Finally,the proposed hybrid recommendation algorithm and two basic algorithms are verified.The performance indexes such as MAE,comparison accuracy,F1-Measured and recall rate are analyzed and compared.The results show that the proposed algorithm has good time efficiency.And accuracy.