I think the main acronym is DIYBI or "Do it Yourself" BI (where BI Is the industry term "Business Intelligence".)
Nelson is actually one of the better guys in the lab, but I do agree that this article was a boring fluff piece.
I think the point is that IBM is extending their BI vision and that since more "unstructured" enterprise data is being captured, there needs to be a way to drill down and find meaning in that data.
I think that IBM is on the right track, however, a lot of the "unstructured" data is industry if not enterprise specific, and trying to create a "standardized toolkit" is about as far as you can go. Really it would be more of a toolkit recognizin g the patterns of the "unstructured" data.
Using the Google-ing of webpages to find information as an example, the tool kit could comprise of some HTML structure knowledge and indexing scheme. It is this form of "intelligence" which is needed.
Of course IBM would need to rethink their extensibility beyond the limited capabilities found in DB2's extenders and apply this DIYBI to IDS first which has a robust enough engine to decrease the time to market and time to value....
But hey! What do I know? I'm just Gumby. ;-)