Shaku Atre asks the question: What Is So Different about Business Intelligence Today? In essence, why the changing nature of computing is driving BI to more of a commodity and useable state. Good points, if you ask me.
"Now we come to the differences between the implementations so far and the BI implementations of today."
She lists:
- Access to computer-based information for most of the employees:
- IT staff needs training in business issues and business terminology:
- IT staff needs training in people communication:
- Presentation of the BI systems:
- BI systems have to follow the scenario of "plug and play" instead of "plug and pray":
- Three-month delivery:
- Web-based EVERYTHING:
- Structured data and unstructured data:
- Mobile computing:
- Tight economy:
The most important changes are really around things we've been discussing in this blog, namely quick delivery of BI systems, unheard of in the past. Moreover, the use of the Web and mobile computing platforms in order to drive the pervasiveness of BI. This includes using cloud computing and the access to Internet-delivered data, that we're able to combine with enterprise data to make better sense of things.
The largest change that I've observed is access and utilization of both structured and unstructured data for use within BI systems. Web data, for instance, once thought useless unless dragged into a relational database can now be accessed and leveraged as easily as structured information, even with semantics bound to that data. Indeed, any information that exists electronically and sometimes not, is usable data these days. This is the core revolution of BI.
As usual, and I've been following her for 25 years, Shaku Atre provides a recitation of the obvious as if she had just discovered it.
BI will evolve into something truly useful when and only when its models and personalization capabilities match the information needs of the people it is meant to serve. Active visualization is one hopeful area. Semantic extraction is another.
-Neil Raden
The models and personalization capability is not moving an inch. Neither there is any breakthrough beyond dimensional modeling which is now coming to its limits. It was a good idea though but we need to move forward. The personalization is a bigger problem and current limitations of RDBMS is the main hurdle. We need a a new data integrity and integration/selection paradigm. Since both these things are not going to happen anytime soon, these people will continue to write about cost cutting and reporting delivered on a mobile phone
Nauman
On the web-based EVERYTHING note, I think another revolutionary move that's occurring with BI is that social networks are taking over. I didn't realize how valuable something like LinkedIn could be when integrated with your business.