数据挖掘技术在档案信息管理中的应用研究

发布时间：2018-06-18 08:19

本文选题：档案管理 + 数据仓库　；参考：《天津大学》2016年硕士论文

【摘要】：数据挖掘是指从大量数据中通过算法搜索隐藏于其中知识的过程,作为一种知识发现技术它已经在很多数据密集型行业得到广泛应用。档案是指人们在各项社会活动中直接形成的各种形式的具有保存价值的原始记录,在我国档案界多年的计算机管理中形成了大量的电子数据,但这些数据的利用还比较初级,只是基于数据库层面的查询、统计、分析等。针对这种局面,本文尝试将数据仓库和数据挖掘技术应用于档案信息管理工作,通过分析档案利用数据库和档案归档数据库,揭示档案与利用者之间、档案与各归档单位之间以及档案之间的关系,从而在更深层次上发挥档案数据在新时代的信息价值。文章主要从档案数据仓库设计、数据ETL、多维数据模型的建立、数据挖掘算法选择、数据挖掘模型应用等方面论述了将数据挖掘技术与档案工作相结合的具体实践,实现了档案数据挖掘的基本过程。首先对构建数据仓库进行详细的需求分析,确定主题和用户,根据已有的档案业务数据库确定数据仓库的维度表和事实表,并最终采用星型模式创建档案数据仓库;然后通过数据ETL将档案原始数据源经过抽取、转换,加载进入数据仓库,这个过程主要是对档案数据进行预处理和清理工作;接着再利用OLAP工具对档案多维数据集进行多角度和多视图的查询,以更好的理解档案数据;最后在档案数据仓库的基础上,根据数据挖掘的需求分析,利用SQL Server Analysis Services工具并选取合适的数据挖掘算法对档案数据进行分析处理,并对结果进行分析。全文的核心为研究数据挖掘技术在档案信息管理中的应用实践。
[Abstract]:Data mining refers to the process of searching hidden knowledge from a large amount of data by algorithm. As a knowledge discovery technology, it has been widely used in many data-intensive industries. Archives refer to various forms of original records with preservation value which are directly formed by people in various social activities. A large number of electronic data have been formed in the computer management of archival circles in China for many years, but the utilization of these data is still relatively primary. Only based on database level query, statistics, analysis and so on. In view of this situation, this paper tries to apply data warehouse and data mining technology to archives information management, through analyzing archives use database and archival archiving database, reveals the relationship between archives and users. The relationship between archives and archival units and between archives, thus exerting the information value of archival data in the new era on a deeper level. This paper mainly discusses the concrete practice of combining data mining technology with archival work from the aspects of file data warehouse design, data ETL, multidimensional data model establishment, data mining algorithm selection, data mining model application and so on. The basic process of file data mining is realized. Firstly, the requirement of constructing data warehouse is analyzed in detail, the subject and user are determined, the dimension table and fact table of data warehouse are determined according to the existing archival business database, and finally the file data warehouse is created by star pattern. Then the original data source is extracted, converted and loaded into the data warehouse by ETL. The process is mainly to preprocess and clean the file data. Then we use the OLAP tool to query the multi-angle and multi-view of the archive cube in order to better understand the archival data. Finally, on the basis of the archival data warehouse, according to the needs of data mining analysis, Using SQL Server Analysis Services tools and selecting appropriate data mining algorithm to analyze and process the file data, and analyze the results. The core of this paper is to study the application of data mining technology in archives information management.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：G270.7;TP311.13

【相似文献】