Data Mining Technology In Logistics Distribution System

During the research of Data Mining Technology in Logistics Distribution System, a data mining subsystem is needed to support the purchase management function in the distribution center of the original LIMS system. First, a data export module has been designed and implemented in order to obtain historical data from the LIMS system and save as XML documents. And then a data import module is necessary to perform the data mining on the exported data. It will import the previous exported data to the data mining subsystem and save it in related database for future use. The most important is the data mining module which is the core subsystem to realize the data mining action with related technology. It includes the core algorithms of data mining and the implementation of association rules. Finally, in order to provide LIMS system the final data and finish the seamless integration without modification of original system, I use the Aspect-oriented programming to add the final data as an aspect into the Distribution Center modules of LIMS system and provide to the purchase management function.Since data mining technology in various industry is being used more and more widely, the relevant industries realize and experience that the combine application of data mining technology and related management systems will be very helpful for their management and business decision analysis. This technology is still in a initial stage of development and needs continuous improvement and development. It needs more professionals to research and design a more efficient association rules and algorithm for mining base on the basic theory for better use. At present, there are a lot of systems using data mining techniques in the traditional and emerging industries, such as e-commerce, logistics, communications, pharmaceuticals, supermarkets, banks and so on. Currently, the technology of data mining is not fully mature. People are engaged in continuing research. Just during such constant process of exploration and research, this technology is being gradually extended. I believe that one day the data mining technology will bring more business opportunities, business models, thinking, and development space to the real world and virtual world in various industries.When people feels the convenience provided by data mining, they will raise more requirement to the research and application of data mining. They will consider deeply on technological innovation and development. The technology itself has the needs of updating as well. Data mining will certainly be improved then. Therefore,during the application of data mining in various industries, there will be various systems based on data mining technology. These systems will work with various sectors in the integration of business details. Therefore, data mining will be a technology which includes technology, human resources, capital-intensive and continuous development and innovation.This thesis discusses the application of data mining technology in logistics and distribution industry. Logistics and distribution industry has become an important industry for people's production and life with the development of electronic commerce, trade, domestic trade., In the traditional logistics information system, organizations purchasement only check the lack of stock base on orders and inventory. It is difficult to estimate the potential needs or the regular needs of future customer. It cannot provide the forecast. Meanwhile, it is also difficult ensure their own product and storage management are set up based on the valid data.These conditions will easily cause the companies cannot respond quickly and provide more personalized service according to the market changes and customer. They should think what customers are thinking, anxious what customers anxious. With the full understanding of themselves and customers, they can provide customers with comprehensive, effective and thoughtful service more fast and more accurately.Due to the needs of the development of the logistics industry and the needs of managing the international marketing, it is the time for data mining technology to apply in the logistics information industry. It also makes it possible that enterprises can respond quickly to meet market needs. The data mining has a bright future in application.This thesis is elaborated the process of design and implementation of LIMS data mining subsystem. First, it reviews the development process of LIMS, introducts generally LIMS system, analyzes on the basic characteristics and disadvantage of the LIMS system, describes the current situation of the logistics management system, and describes the cause of data mining technology application in LIMS systems. Later, I explain the relevant technologies involved in this thesis. I introduce the general overview of data mining, and analyze of the basic profiles of Struts framework and the application target of Struts framework in data mining subsystems. The framework is a good choice for the system implement. Spring framework and Hibernate framework are popular open source framework in current system development. The excellent design and implementation makes the maintenance and management of system middle layer and persistence layer components easy. It also provides more convenient and low-cost solution for the system upgrade and expansion. Then I describes on the XML technology and application accordingly.In this thesis, I discusses the important role of related technologies and applications in data mining subsystem. After the description of the related technology, I show the architecture of the subsystem component. Then, through the requirement analysis of the subsystem, I specify the subsystem design objectives. I describe how to combine the different technologies in the target system as well. I tried to export the data from original LIMS system using the new subsystem and save it as XML document, then import the data using data import module into the data mining subsystem. Finally, I send the finished data back to LIMS system. The implement process is based on the standard software life cycle which goes through requirement analyze, general design, detail design, code implement and unite testing and system integration testing.During the design and implementation process of subsystem, the popular technical architecture and application makes it easy for subsystems to achieve the smooth integration with the original system. It makes developers to improve the efficiency, shorten the development cycle and enhance the system's maintainability and scalability. Subsystem allows the system can eventually be achieved on the parent's data collection and export, to export data to XML format for preservation. Queries can be combined with a variety of conditions. Also it achieved the export data to be imported into the data mining subsystem. This process involves the translation of XML document types and database data format. With mature data mining algorithms and practice, it achieve the data mining of purchase data, and successfully applied back to LIMS systems.The implementation of the system provides huge data support of the management decision analysis of the logistics industry. It is the base of future design on the more complex analysis of logistics industry In the logistics industry, data mining technology has not be used widely and only has few functions. The research in this thesis on the data mining subsystem of logistics information management system has a profound significance and value to the information management in logistics and it will bring more potential opportunities to the logistics industry.