Science meets DWH and BI.....and vice versa.

I have just finished reading several papers that describe the history of DSS (Decision Support Systems)-research [1,2,3,4,5]. Most of these papers are so-called meta-analysis studies.

First of all; I am doing research in the field of Information Management. Decision Support Systems (DSS) are part of that field of research. For this blog post I will use the categorisation of Arnott [4,5] within DSS:
1 - Personal DSS (PDSS)
2- Group Support Systems
3- Enterprise reporting and analysis (includes BI)
4- Data Warehouse
5- Intelligent DSS
6- Knowledge management-based DSS
7- Negotiation Support Systems
8- others

Interesting quote from Arnott [4];
'PDDS remains an important aspect of the IT-based management support in contemporary practice. Modern PDSS can source data from data warehouses and deploy powerfull modelling approaches from management science/operations research. The current industry term for the later class of PDSS is 'Analytics'.'

I like this quote because it de-hypes the phrase 'analytics' a bit.....a bit more grounding which is needed!!

Three things strike me when I read the categorisation of these DSS types:
1- Arnott puts Business Intelligence in category three. Let me quote Arnott [4] again:
'Business Intelligence is a poorly defined terms and it's industry origin means that different software vendors and consulting organizations have defined it to suit their products; some even use 'BI' for the entire range of DSS approaches. We use Business Intelligence as the contemporary term for both model-oriented and data-oriented DSS that focus on management reporting, that is, BI is a contemporary term for EIS.'

I very much agree with Arnott and would like to go that extra mile here; Business Intelligence, as a term, is dead. I would like to summarize the above quote into the fact that Business Intelligence in the idustry is badly grounded in any form of theory. For the sake of this post I will use the term Business Intelligence as defined by Arnott.

2- Arnott distinguishes between data warehousing and business intelligence as two different types of DSS.

Again; something I teach (and preach) for a long time. It will make a more rigor discussion possible if we would all seperate DWH from BI and vice versa.

3- Data warehousing is still being pushed into the decision support category of IS research. I see a DWH more as a prerequisite - in some cases - for decision support. A data warehouse by itself can not support decision making.....and it should be noted that modern day data warehouse not only should be able to support decision making, but also other processes.

I won't go deep into the articles, but let me give you some conclusions:
- There is an extreme low proportion of scientific papers on DSS types 3 and 4
- There is huge indication that DSS Types 3 and 4 are largely responsible for commercial expenditures in DSS.
- In general; the DSS research field is facing a crisis of relevance. More than half of all DSS research was assessed as having low or no practial relevance
- Only type 3 and 4 have reasonable relevance scores
- But...less than 10% of all DSS papers/research are in this area and data warehousing is only at a staggering 1.3%
- In most of the papers there is no reference to any judgement-or decision making methodology.

What do you guys think of this? Aint it weird that companies are investing (generalizing a bit here) hugely in data warehousing and Business Intelligence and their is virtually no scientific attention to these DSS types?  Data warehousing and Business Intelligene are very poorly grounded in science and yet companies seem to throw billions at it.....

Two things;
- The scientific community needs to wake up. They need to forget about the low relevance DSS Types and focus on the ones that seem to matter.....
- Practitioners in data warehousing and Business Intelligence need to strenghten the theoretical foundations a hell of a lot.

Science meets DWH and BI.....and vice versa. I got my work cut out for the years to come.

Edit may 27^th: Partner in crime - Wouter van Aerle - made a good post on his blog regarding this post

[1] Harvard Business Review - J.Dearden - MIS is a mirage - 1970
[2] Decision Support Systems - S.EOM, S.Lee - Leading US universities and most influential contributors in decision support systems research (1971 -1989)
[3] Decision Support Systems - I.Benbasat, B.Nault - An evaluation of Empirical research in Managerial Support Systems - 1990
[4] Journal of Information Technology - D.Arnott, G.Pervan - A critical analysis of decision support systems research - 2005
[5] Decision Support Systems - D.Arnott, G.Pervan - Eight key issues for the decision support systems discipline.

Comments

[Wouter van Aerle]

Good post. I've added a comment on my own blog.

Cheers,

Wouter

[Sjoerd Venema]

Food for thought ;) Inspires me to read more on this.