Data Warehouse Structure In NHS Facilities And OLAP Solution

Business Objects is a famous Business Intelligence system which made by a software company also named Business Objects. User could use Business Objects to design and build a total new data schema base on source database without changing the structure of the source database in physical layer. This schema made by Business Objects called Universe, and all the business reports could be made from it.National Health Service (NHS) is the national medical system of UK, which build after the Second World War, has a nearly 60 years history. NHS electronification has been started from computer popularization time. Because the quantity of medical facilities is so huge, nearly every terminal system and database of facilities is different from each other. And the database of some old hospitals is Enscribe database, which is different from normal SQL one, but all tables in it are files, and it doesn't support SQL query. For some new hospital, there a total SQL standard database. So, NHS wants to build a unique terminal system and database. This project started from 2003, anticipation is 10 years. Meanwhile, there is a project named Operational Reports Project which takes the charge of supporting reports for all Trusts in NHS.But, at the very beginning of Operational Reports Project, NHS has made a fatal false: They just build a new database from Enscribe database and SQL database to a HP Non-stop Tandem. It is just a database but not a data warehouse, which totally can not meet the request of a data warehouse. So, we cannot make effective reports from this database. And NHS does not want to build a new data warehouse with any more money.This paper is showing a method using Business Objects Business Intelligence system to build series of Universes from NHS HP Non-stop Tandem without hurting the original database schema. And all the reports will be made from these Universes. Universes in this solution are expansible and easy for maintenance. At the first step of design, we have thought about the second round development for every Universe. This paper talks about the key part of Universes and Reports development. And also gave every solution according to some actual problems during developing.First Chapter talks about the concept and background information of data warehouse, such as OLTP, OLAP and Business Intelligence, and gives the foresight of their developing. Data warehouse makes people see the trend and orderliness in the ocean of data. With the developing of database and the growing of e-business, the quantity of data is very fast. In this chapter, I have described the summary for the development of data warehouse nowadays in my opinion. And have shown my own view of some popular words such as Data Integration and Business Intelligence. At the end of first chapter, we also could see someinformation of demographic and geography in UK.Second chapter described the main architecture and solution of Operational Reports Project. Firstly, we will see the process routine from terminal system to back database. And I will talk about how to build a Business Objects system in this environment. In the next two sections gives the dimension model of Universes and Reports. During this chapter we could see something different between Universe and data warehouse, also we could see they are something in common. I have shown the defects which came from the difference, and the fixing method for it.In chapter 3 and chapter 4,1 have set "Others" Module as an example, which I was in charge. This module includes 12 reports. We will see some difficulties during the developing of "Others" Universe, and solutions for them. "Others" Universe belongs to QMS Trust which is a new Trust of NHS. This Universe is big;the development for it is an expansible process. These 12 reports is a representative collection, in chapter 4, I also took one report as an instance to show the steps of making reports. Meanwhile, because HP Non-stop Tandem has defects which can not parse the right outer join. It makes the development of reports have to avoid right outer join. We will see the solution in chapter 4.At the end of this paper, I have made a summary of Operational Reports Project. Raise some main issues of this project. And because of its long term, I also give my own opinion for its development.