Design And Implementation Of A Personalized Recommendation System For Domain Of Logistics Distribution

With the rapid increase of the economy and the e-commerce, logistics distribution has got great promotion and development. Nowadays, the quantity of distribution service plans provided by logistics is huge. However, the quality differs much. Customers have different demands of the distribution price, distribution speed, safety, experience, assurance and so on. Therefore they often suffer from picking a right plan that fits their personal demands and has the quality assurance. How to improve the efficiency, service quality and reduce service costs and to meet customer personalized demand has become a hot issue in the field of logistics research. Targeting on the issue of how to build a recommendation system and then to recommend distribution service plan to meet customer personalized demand, this paper mainly studies the following aspects:(1) Targeting on the issue of the assurance of logistics distribution service quality, the methods of evaluating quality of distribution service plan are studied. A method of the evaluation of distribution service quality based on information system is proposed in this paper. The service evaluation index system is established based on the current studies and the weight of each index in the evaluation index system is calculated based on Analytic Hierarchy Process and causal relation method of System Dynamics. The comprehensive evaluation value of the distribution service quality is calculated based on liner weights method.(2) Targeting on the issue of the cold start in the recommend system, a recommendation method based on service quality is proposed based on the distribution service evaluation results. The content based recommendation method is applied in the system based on the evaluation index system got from the distribution service evaluation. The effectiveness of these two methods are validated based on simulation data in an experiment and the advantages and disadvantages are analyzed consequently.(3) Targeting on issues of data sparsity and recommendation efficiency in the collaborative filtering method which is applied most widely, a collaborative filtering method based on user’s interests on service attributes is proposed and implemented. An experiment is carried out comparing with the traditional collaborative filtering method based on the simulation data. The results of the experiment prove the efficiency on these two issues.(4) The status of users in the recommendation system is analyzed. And for each status, an effective recommendation method is specified after the experimental analysis on the applicability of the content based recommendation method, the service quality based recommendation method and the collaborative filtering method based on user’s interests on service attributes.(5) The requirement of the personal recommendation system for the domain of logistics distribution is analyzed. And then the system module design and database design are completed after the module division of the system. Finally the implementation techniques are introduced and the system are implemented and tested.