Clustering Algorithm And Its Application In A New Management System For The Small And Medium Enterprises

With the rapid development of information technology and the promotion of"Internet+" concept,the management systems play an increasingly significant role in the operation of Small and Medium Enterprises(SME).However,information has been lagging behind due to limited resources available to SME.Hence,ways to utilize technological innovation to enhance the efficiency of SME has become a topic with growing attention.Through communication with enterprises,requirements for SME have been analyzed and a newly designed management system is proposed.This system integrates core business function,advanced technology method,data mining algorithm and other resources to support SME in their decision-making process.The finding from this project is able to facilitate enterprises to incorporate a convenient,efficient,and affordable management system.The main research results include:(1)A lightweight technical architecture designed for the needs of SME.Based on the MVC design pattern,the technical architecture utilizes a mode of development that highly combines the JFinal technology framework and the Enterprise Wechat Accounts.The architecture ensures the prompt development and application of the system with the premise of security and stability of the management system.In addition,four core functional modules according to the business services of SME-employee management,income management,supplier management,and Enterprise Wechat Account service-have been proposed.This inclusion of services can safeguard the integrity of the new system;(2)A K-means algorithm based on optimal attribute weights that target the absence of scientific classification management of suppliers for SME.This algorithm improves the traditional K-means method by constructing an attribute weight optimization model,reference to the Lagrangian function to calculate the optimal weight value of each dimension.The automated attribution of datasets can objectively reflect data distribution on Euclidean space and thus reflect the actual significance of each attribute.Extensive experiments demonstrate that the proposed algorithm has outstanding performance;(3)A new enterprise management system adapted by several SME designed with the method and improved algorithm described in(1)and(2),respectively.This system is constructed by integrating the newly designed technical architecture and the improved K-means algorithm,while utilizing the core functional modules as described in(1).The function and performance of the system are able to meet the needs of SME and therefore greatly enhance the efficiency of its management and implementation in enterprises.The system has passed relevant enterprise acceptance tests and is being operated on-line with smooth progress.