The DAMA Guide to the Data Management Body of Knowledge1 (DM-BOK) sets out to provide a compilation of principles and best practices and to provide practitioners with a framework to manage data and to ‘mature’ their information infrastructure. These laudable aims can be judged at two levels. First by how well the book achieves its stated aim and second, how appropriate the isolation of enterprise data management is as a discipline—given that there is in every enterprise, a constellation of domain specialists, database managers, IT hardware and software experts who, to a greater or lesser extent already occupy the ‘space’ delineated by the DM-BOK.
We were disappointed to find that the Guide does not contain a glossary or definitions—these were issues in a previous publication. Also many interesting topics in the index—for instance ‘geospatial meta data standards’ are one liner links to an external website. So to get the full benefit of DM-BOK you need to acquire a) the Dictionary and b) a few hundred reference works. This would not be too bad if there was any indication as to what the really essential works actually were. Another irritation is that, instead of a chapter on ‘master data’ or ‘data quality’ there are chapters on ‘master data management’ and ‘data quality management.’ This allows DM-BOK’s authors to speak from the management high ground rather than addressing how things get done.
It is not all bad though. The chapter on data quality introduces the Demming cycle and gets into tools and tricks for cleaning up names—although it would have been nicer to name some of the tools actually in use! Data ‘entropy’ gets a good treatment as does the idea that it is better to fix data upstream rather than before it gets trashed—building quality into the data resource. There is also advocacy for a single data architecture as key to quality in enterprise data.
One problem with the Guide is that, because it is building a ‘profession,’ it is too abstract and offers too little in the way of concrete examples. This combines with a tendency to slip into database jargon and hampers understanding. To give an example. A ‘foreign key’ is described as a ‘an attribute that provides another link to an entity.’ Rather opaque when compared with Wikipedia’s ‘a column that refers to a (..) column in another table.’ In a similar vein, DM-BOK ploughs through one-liner definitions of normal forms—up to number six—without offering any real insight as to what is going on. One can imagine DM tyros having to rote learn this stuff, perhaps chanting it out like mid 20th century schoolchildren learning math (well at least they did learn it then!).
But perhaps more importantly than misgivings that one might have about the DM-BOK guide is the feeling that it fails to make as good a case as it could have for the existence of a DM professional as for instance the accounting profession. Read this month’s editorial for more on this.
DM-BOK’s level of abstraction would imply that a DM professional could switch between say E&P and banking. To take an extreme example, an SME who understands geodetics is unlikely to be severely tested by the content of DM-BOK. But he or she might pick up some useful jargon and understanding of the vast overlapping collection of technologies and solutions that make the field. A manager on the other hand may get the impression that this is easier than it really is.
1 Data Management Association Guide to the Data Management Body of Knowledge. ISBN 978-1-935504-02-3, $74.95.
This article originally appeared in Oil IT Journal 2010 Issue # 6.
For more information or to comment on this topic email here.