【作者】 程扬；

【导师】 李晓；

【作者基本信息】 新疆大学 ， 计算机应用技术， 2006， 硕士

【摘要】 近年来,随着医院制度和社会保险制度改革的深入,一般的大中型医院都相继建立了自己的医院管理信息系统(HIS)。医院对信息系统的管理越来越重要。然而,随着HIS的应用和不断发展,HIS数据库中的数据也在日益的膨胀,如何充分分析利用医院现存的这些大量宝贵的信息资源,为医院的管理决策服务,使医院在医疗市场竞争日益激烈的环境下,保持自己的核心竞争力,不断发展,基于HIS的数据仓库技术和数据挖掘技术应运而生。基于这种现状,本文在通过研究现有HIS的基础上,提出了建立基于医院管理信息系统的数据仓库,以及进行数据挖掘分析来提高医院管理决策水平的一种解决方案。通过数据仓库技术,根据实际需求,从医院信息海量数据库中分析、提取、确立主题,进行有效的数据组织,来构建数据仓库模型。在进行数据清理和数据转换后,实现对数据仓库的数据装载。从而,对创建好的数据仓库就可以进行SQL查询,报表统计,OLAP数据分析以及数据挖掘等方面的应用,来有效地服务于医院的全方位管理决策。本文运用了Microsoft SQL Server 2000提出数据仓库与数据挖掘技术的解决方案。Microsoft公司的SQL Server 2000已经在性能和可扩展性方面确立了世界领先的地位,是一套完全的数据库和数据分析解决方案。Analysis Services提供了OLAP和数据挖掘的功能,将OLAP功能集成到Microsoft SQL Server中,提供可扩充的基于COM的OLAP接口。本文使用OLAP和数据挖掘技术,通过分析数据仓库中的数据,对多维数据集进行构建,从多层次、多角度对HIS数据仓库门诊部和住院部信息进行OLAP数据挖掘和分析,并调用PiovitTable进行了分析结果的展示。同时,提出了对关联规则挖掘算法的一种改进,并且给出具体算法实现,该方法提高了实际应用中关联规则挖掘的效率。具体来说,本文做了以下工作:通过对源数据库中的数据进行大量数据分析和数据预处理工作,提取出构建数据仓库的主题,提出了基于HIS的数据仓库的模型。提出了使用Microsoft SQL Server 2000 Analysis Services进行OLAP与数据挖掘,并对数据仓库数据创建多维数据集,运用OLAP技术进行多维数据分析和数据挖掘。研究和探讨了关联规则挖掘算法Apriori算法,以及其运用于数据挖掘中,提高算法运行效率的几种方法。并且研究利用Apriori算法,提出了一种有效的改进方法,对现有的关联规则挖掘算法进行了改进和扩展。更多还原

【Abstract】 With the development of the hospital system and the reform of the social insurance system, a great many of hospitals have created their own hospital management information system (HIS). Hospital management information systems are increasingly important. However, with its applications and development, its database is growing inflation, the analysis of how to fully utilize the existing hospital valuable information resources for hospital management decision making services and in the medical market increasingly competitive environment maintaining their core competitiveness, continuous development, so, the technology of DataWarehouse and DataMining based on HIS come out.In this situation, through researching on HIS existing , thesis proposes the establishment of a Data Warehouse based on hospital information management system and data analysis of Data Mining to improve hospital management of the decision-making level. Through the technology of data warehouse, based on actual demand, to analysis, extraction, established themes, effective data organization from enormous hospital database, to build a data warehouse model. After Data Cleaning and Data Transforming, the Data Warehouse comes to Data Loading. Thus, the establishment of DataWarehouse can hold SQL Queries, statistical Reports, OLAP data analysis and Data Mining, to effectively serve the hospital management decision-making.The thesis makes use of Microsoft SQL Server 2000 technology of DataWarehouse and DataMining to put forward solutions. Microsoft SQL Server 2000 in the performance and expansion have established world-leading position,is a competitive database and data analysis solutions. Analysis Services provided OLAP and DataMining functions ,and made OLAP become the integration of Microsoft SQL Server, provided expansion of OLAP-based the Com interface. The thesis makes use of OLAP and DataMining techniques, analysis of data through the data warehouse to carry out a multidimensional datasets from the multi-level and multi-angle,and analysis of data warehouse of hospital out-patient department and in-patient departments of information by OLAP and DataMining analysis, and the results of the analysis available PiovitTable display. At the same time, the thesis puts forward an effective Association Rules mining method to make an improvement, and gives specific algorithms. This method enhances the practical application of efficiency of Association Rules. Specifically, this article has done the following jobs:Through a great deal of data analysis and pre data processing jobs from the Source database, put forward the theme of constructing the data warehouse and his made based on the data warehouse models.Made use of Microsoft SQL Server 2000 Analysis Services for OLAP and DataMining and the creation of multidimensional datasets by data warehouse,and used OLAP technology for multidimensional data analysis and DataMining.Researching and exploring the Association Rules mining algorithms and Apriori algorithm ,and its application to Data Mining of several ways to improve the efficiency of the algorithm operation. Apriori algorithms and the use of research, made an effective way to improve the existing rules related to the improvement and expansion of the Association Rules mining algorithms.The organizational structure of the article is as following更多还原