聚类算法及其在新型中小型企业管理系统中的实现

发布时间：2019-07-08 20:11

【摘要】：随着信息技术的飞速发展和“互联网+”思想的推广,企业管理系统在中小型企业经营上扮演着越来越重要的角色,由于中小型企业的资源有限导致信息化水平落后,因此如何利用技术创新提升中小型企业的效率成为研究热点。本文通过与企业的交流,分析需求,提出并实现一个集核心业务功能、先进技术方法、数据挖掘算法等资源为一体的新型中小型企业管理系统,帮助中小型企业低成本地引入一个便捷高效的管理系统,辅助企业决策。本文的主要研究成果如下:(1)针对中小型企业的实际需求,设计了一种轻量级的技术架构。该架构基于MVC设计模式,采用JFinal技术框架和微信企业号高度结合的开发模式,保证了文中企业管理系统在具有安全性、稳定性的前提下,既能快速开发又能快速应用。除此之外,针对业务需求,设计出四个核心功能模块:员工管理、收入管理、供应商管理和微信企业号服务,确保新型系统的完整性;(2)针对中小型企业缺乏对供应商科学性分类管理的需求,提出了一个基于自适应属性权重优化的K-means算法。该方法在传统K-means算法的基础上,通过构建属性权重优化模型,引用拉格朗日乘数法自动计算数据各维属性的权重最优值,客观真实地反映出数据集在欧氏空间中的分布情况,以此反馈不同属性的重要性。实验验证了所提算法的有效性;(3)依据上述提出的设计方法与改进算法,实现了轻量级的新型中小型企业管理系统。该系统基于本文设计的技术架构,集成了本文提出的算法,并依据设计要求实现了系统应具有的核心管理模块,在功能和性能上均已满足中小型企业的需求。本文所依托的系统项目可以大幅度提升中小型企业的管理和执行效率,已通过相关中小型企业的验收,成功上线运营且运行平稳。
文内图片：

图片说明：图３－１邋ＪＦｉｎａｌ框架结构图逡逑Ｆｉｇｕｒｅ邋３－１邋ＪＦｉｎａｌ邋ｆｒａｍｅ邋ｓｔｒｕｃｔｕｒｅ逡逑
[Abstract]:With the rapid development of information technology and the promotion of the idea of "Internet", enterprise management system plays a more and more important role in the management of small and medium-sized enterprises. because of the limited resources of small and medium-sized enterprises, how to use technological innovation to improve the efficiency of small and medium-sized enterprises has become a research focus. Based on the communication and analysis of requirements with enterprises, this paper puts forward and implements a new management system of small and medium-sized enterprises, which integrates core business functions, advanced technology methods, data mining algorithms and other resources, which helps small and medium-sized enterprises to introduce a convenient and efficient management system at low cost to assist enterprises in decision-making. The main research results of this paper are as follows: (1) to meet the actual needs of small and medium-sized enterprises, a lightweight technical architecture is designed. The architecture is based on MVC design pattern, and adopts the development mode of JFinal technical framework and WeChat enterprise number, which ensures that the enterprise management system in this paper can be developed and applied rapidly under the premise of security and stability. In addition, according to the business requirements, four core functional modules are designed: employee management, income management, supplier management and WeChat enterprise number service to ensure the integrity of the new system; (2) aiming at the lack of demand for scientific classification management of suppliers in small and medium-sized enterprises, a K-means algorithm based on adaptive attribute weight optimization is proposed. On the basis of the traditional K-means algorithm, by constructing the attribute weight optimization model, the Lagrangian multiplier method is used to automatically calculate the optimal weight value of each dimension attribute of the data, which objectively and truly reflects the distribution of the data set in Euclidean space, so as to feedback the importance of different attributes. Experiments verify the effectiveness of the proposed algorithm. (3) according to the proposed design method and improved algorithm, a new lightweight small and medium-sized enterprise management system is implemented. Based on the technical architecture designed in this paper, the system integrates the algorithm proposed in this paper, and realizes the core management module of the system according to the design requirements, which has met the needs of small and medium-sized enterprises in terms of function and performance. The system project relied on in this paper can greatly improve the management and implementation efficiency of small and medium-sized enterprises, and has passed the acceptance of relevant small and medium-sized enterprises, and has been successfully operated online and runs smoothly.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP311.52;TP311.13

【参考文献】