工资智能决策支持系统的研究与实现

【作者】 黄鹤远；

【导师】 汤庸；

【作者基本信息】 广东工业大学 ， 计算机应用技术， 2000， 硕士

【摘要】 随着计算机应用的发展，人们已不再满足于数据库向他们提供的事实、报表。特别对决策者而言，他们不仅需要各种数据，更需要的是根据数据进行科学的统计分析并提出合理的决策方案，供有关人员参考。 决策支持系统（DSS）是用于处理半结构与非结构决策问题的计算机软件系统，传统DSS中一般仅通过数学模型等方法来辅助决策，但由于现实生活中的很多问题难以用纯粹的数学模型加以表示，因而具有较大的局限性。若能够将数学模型与知识推理相结合，系统的适应面将大大拓宽。智能决策支持系统（IDSS）正是适应这种要求，将人工智能、数据库与决策支持等融为一体，更为理想的满足了实际决策工作中的需要。 我们与广东省人事厅合作开发的工资智能决策支持系统，是一个智能决策支持系统在实际的软件开发中的具体应用。其目标就是将各单位的人事部门从繁琐的工资调整工作中解放出来，不再陷于成堆的工资政策中。同时也可以避免许多人为造成的错误，使工资管理规范化。另外该项目也积极配合广东省财政统一工资发放的要求，对工资信息来源的正确性作出保障。 本文描述了工资智能决策支持系统的体系结构和思想。该系统是一个三库结构的系统，包括规则库、方法库和数据库。与四库结构相比，它具有独立性较好，易于实现的特点。更多还原

【Abstract】 According the development of computer application~, people can not settle for the details and reports provided by database. Considering decision-makers, they need not only all kinds of data, but also the recommendation for decision-making from the statistic and analysis of these data. DSS is the computer software system to process semi-structured and non-structured problems. Traditional DSS assist us to make decision by using mathematical model, however, in the reality, a lot of problems could hardly be represented with only mathematical model. If we can integrate mathematical model with knowledge reasoning, the system will become more adaptable. According to the demand, IDSS amalgamates with artificial intelligence, database, decision support, and it is more suitable for actual decision-making. The SIIDSS developed under our cooperation with Personnel Department of Guangdong Province and us is an application of IDSS. It抯 goal is to disengage personnel of every organs from drab work of salary regulation and enable them break away from a great deal of policies of salary. At the same time, by making use of the SIDSS, every organ can avoid a lot of man-made mistakes and standardize the management of salary. In addition this project responds to the demand of unifying putting out salary by treasury department of (iuangdong Province, guarantees the validity of salary. This paper introduces the architecture and the idea of SLDSS. The system include three bases: rule base, method base and data base. Compared with the system of four bases, the modules of our system are more independent, and our system is easier to be implemented更多还原