For many years, business intelligence (BI) professionals have had to learn on the job, from colleagues or consultants when advancing their skills.
A masters programme at the University of Dundee, now in its second year, is providing an academic outlet.
“Doing this course has made me realise how common a lot of BI technology is. It’s just different implementations”, said Chris Hillman, a student in Dundee’s masters programme in business intelligence, and a business intelligence technical director at market research firm Symphony IRI, in Bracknell.
Mark Whitehorn, senior lecturer at the university, teaches the course and said the core question it answers is, “How do you make a BI project work?” Subsequently, there is a big focus on gathering user requirements.
- BI systems – introduction and overview
- Relational database design
- Dimensional database design
- Designing BI systems
- ETL (extract, transform and load) – theory and practice
- MDX (multidimensional eXpressions)
It is also available as a part-time course. Although students are expected to have a computer science degree, they can be graduates of other disciplines or have relevant professional experience.
Whitehorn confirmed that students are mostly in their 30s and 40s, and mostly from the UK, though some are from continental Europe and Africa.
“Almost all have an IT background – leaning towards the business analyst side”, he said. However, an area of development is “finance people who are heavily involved with IT as users”.
Current students are from financial services, education, police, retail and scientific research.
The course features external speakers from the IT industry, but “vendors can’t do a sales pitch, and our students have been trained to be merciless if they do”, Whitehorn said.
Martin Willcox, director of platform and solutions marketing, Teradata Corporation, and David Hobbs-Mallyon, SQL Server product manager from Microsoft, have publicly expressed support for the course.
Whitehorn said he would favour a professional body for business intelligence, of which the Dundee course could be a proponent. He also maintains that the broader academic scientific world tends to do data badly, and that Dundee is aiming to develop a centre of excellence out of the masters programme that would see BI applied to natural and social scientific research. An instance of this, said Whitehorn, is the School of Computing’s work on proteomics data generated by Professor Angus Lamond’s team at the Wellcome Trust Centre for Gene Regulation and Expression at Dundee.
Hillman said he has found the course “useful in the day to day. Being out and about with our sales staff talking to clients you can talk more easily on a generic level, for example about the [Ralph] Kimball methodology”. He said that the masters programme has educated students more deeply about the role of the BI architect, and offered a rounded formation that is as much about user requirements as it is about technical detail.
“If you took the population that understands relational databases properly and then took out the people that understand hierarchies and multidimensional databases, and then took out the ones that can speak to real people, you are not left with many,” Hillman said. “You have to be enough of a geek to be interested, but not too much of a geek. And you have to love the data.”