Desktop search has become an important component of our everyday work.
With the amount of information explosion, it is only natural that users and enterprises move towards enabling desktop (and enterprise) search for users – subject of course to appropriate security and access controls. BI vendors have moved into this new business space that has opened up and seems to be one of the most promising. While Business Objects had announced support for the Google Search appliance and Google Desktop back in 2006, their most important announcement lately has been the launch of the Business Objects Explorer (formerly known as Polestar) product. More about that in a later post…
Having relocated from the Silicon Valley to Bangalore a year back, I’m now working in an MIS – strategic reporting role. In my role to evangelize the use of BI best practices and tools, one of the foremost is that of universe design. As a matter of fact, I’m currently being involved in formalizing a BI policy around the tools we use most – Oracle, Informatica and SAP Business Objects (along with migration from our legacy BO to the XI platform!) – so a lot of my current work is related to best practices, design guidelines and preparing unit test checklists for my team of developers.
There is a well known adage that if you keep doing the same thing and expect different results, that is a sure sign of idiocy.
In the BI world too, we come across several instances where people take it for granted that the ‘BI tool’ will magically generate insight and spur ‘intelligence’ rather than ‘idiocy’. Yet the very practices of reporting the same measures, or of creating reports for metrics just because they are now made available by the tool, without sparing any ‘intelligence’ into what will generate insight is a major cause of failures of BI.
Not since the late seventies, when Larry Ellison’s Relational Software Inc. (RSI) turned out the first commerically available RDBMS - Oracle, has there been such rapid changing of the rules (read disruption) in the database industry.
With Web 2.0 pushing enterprise adoption, and the ensuing information explosion in the maze of audio, video, data and ever-growing data warehouses, it seems that the conventional relational database systems are growing tired. With estimates of unstructured data being anywhere between 80% to 95% of all business data, and the ever changing requirements imposed by Web 2.0 - storage of pictures, audio and video, the demands being made on conventional RDBMS technology are monstrous. With the load window available being fixed due to availability and uptime requirements, the ever increasing data to be loaded into data warehouses, the bulking-up of the data due to usage of XML based formats, conflicting requirements of SQL and XQuery, the database is also being challenged by the demands of business intelligence.
Recently I worked on an assignment where the Business Objects universe was designed by a business analyst. Nothing wrong with that, as long as we get the correct results in reports or queries on the universe. But it was not to be.
We encountered incorrect results primarily due to two reasons:
Mixing of grains in fact tables by not using proper contexts or the aggregate navigator
Unresolved chasm traps
It was obvious that the business analyst was not trained in data modeling, nor did the universe go through a proper QA cycle where it is reviewed for data modeling errors.
