Thanks for weighing in. My quick response: (1) I like your definition’s inclusion of no advance knowledge of schema or queries, (2) I agree that they must go beyond returning list of links to documents, but feel that saying they must be mash-ups is overspecifying, (3) I like the whole fuzzy angle as a distinguishing characteristic; databases and data-driven application don’t do fuzzy well, (4) it seems artificial to argue that they must including structure and unstructured data sources as well as the web because you can simply argue the web is an unstructured data source. Some people (e.g., reputation management) might really care about it. Others may have no interest whatsoever, but it doesn’t make their application any less search-based.

While I won’t edit it out — and since I’ve just switched blogging tools, I’m not sure I know how– please refrain from the little ads in the last paragraph, otherwise I might have to reject the whole comment which I don’t want to do. Merci / Dave

I would guess that the key distinction for the IDC would be this statement:
“They deliver a purpose-designed user interface tailored to support a particular task or workflow.”

I’d guess that I’m like most people who think of “search platform” as being something like ESP, IDOL, MDEX, etc., etc. — they either provide no U/I at all, or if they do, it’s a simple U/I that just handles general search.

Thoughts?

Avi

First, the core of the SBA concept is based on providing search-like query interactivity that doesn’t require advance knowledge of a specific query language (e.g., SQL) or data structure (e.g., multi-dimensional cubes or data entity relationships). Clearly this opens use to a dramatically broader user community than other technologies that have those prerequisites .

Second, SBAs present information in mash-ups that can be more than just a list of page or document hits. Our customers want this to accommodate specific business processes or consumer activity needs.

Third, SBAs can integrate information using fuzzy logic, semantic meaning and other means beyond simple key-based joins or term matching.

Lastly, that information can be from not only Enterprise structured and unstructured data sources, but also the Web. The win being a richer, more real-time, more nuanced context for analysis and decision-making, be it consumer or business.

Of course, even though we recognized and even put a name to this kind of application early on won’t keep other related technology vendors from recognizing the market opportunity in their own way and joining in. That’s completely understandable. Consider the total revenue of the BI market over the last decade and the size of the data upon which that industry was built.

Now, fast forward to now, and consider the current data generation rate, the majority being unstructured or modestly structured information and then extrapolate out the business value of harnessing that information.

So, this is why Exalead is in this domain.
As part of this commentary, I’d like to THANK OUR CUSTOMERS for collaborating with us in creating some very special and distinctive search-based applications. It is their progressive vision in their respective industries that help us drive our innovation. To our customers, present and future, I say: “please stay tuned there’s more innovation yet to come!”

François Bourdoncle, Founder, Exalead