Congratulations to Arwa Jamjoom for successfully passing her MPhil to PhD transfer
Monday 29 June 2009
Arwa successfully transferred on Wednesday 24th June with her work on "A Methodology for Clinical Data Warehouse". Arwa is supervised by Professor Paul Krause and is a collaborative student studying in Jeddah. She was examined by Dr Matthew Casey with chair Dr Lilian Tang.
Purpose / Objectives: this research proposes a framework for developing a data warehouse focusing on the requirements specification phase. A prototype method is presented, along with a modelling method. The method is being developed and will be evaluated in close co-operation of a leading hospital in Saudi Arabia (KFSH&RC-J). It is expected that the findings of this research can be used to help understand factors contributing to the selection of data warehouse method and to provide techniques to overcome expected challenges and difficulties.
Background: the continuous growth in research and the development in data warehouses have been motivated by the awareness that there is a number of potential benefits of this technology and that the vendors of such systems are offering competitive products. Meanwhile, the process for building a data warehouse is complex, expensive, and time consuming. Data warehouse projects are inherently risky with significant causes and sources of failure, in addition to the difficulties of any technology project. Data warehousing faces a number of challenges in terms of cultural perspectives, proper selection of the development approach, and standardization. As the business requirement analysis phase is one of the most important factors in data warehousing success, there is a need for specifying and gathering the right information in a more comprehensive, effective, and formal method, anticipating future requirements with less effort, and exploring the ways to stimulate end users to be more creative about their needs.
Method / Design: an intensive literature review has been conducted covering: (1) review and analysis of the leading data warehouse development methodologies and most common approaches, (2) survey of ten real life examples on developing data warehouse in healthcare industries, and (3) study of related work in data warehouse development methodologies. In addition, at KFSH&RC-J, a total of 18 professionals have been interviewed: they gave 33 interviews lasting over 30 hours; moreover, three training workshops have been attended lasting a total of 8 hours to gain familiarity with the systems and, more importantly, with the end-users of the systems. As a result, a proposed methodology supported by guidelines and selection criteria of data warehouse method will be presented.
Result / Conclusion: Current research on data warehouse development methodologies are either cases to develop a standard platform, cases to improve a single method approach, or cases that integrate more than one approach together. Moreover, different kinds of techniques have been adopted such as use cases, ADAPT notation, and the Goal/Question/Metric paradigm to improve the business requirements phase. Yet still it face a lot of difficulties. However, adopting concepts from Agile development processes may bring substantial benefits into the data warehouse such as customer satisfaction, late changing requirements, early and continuous delivery of the project.