Unstructured Information, ECM and Enterprise Information Management
By billy.cripe on May 19, 2008
I've been thinking a lot lately about synthesis. I've been thinking a lot lately about composite applications and enterprise mashups (aside: check out mashup gaming at: PMOG, hat tip: Jake). Though Jake doesn't like the term "Enterprise 2.0" and I do, I think that there are some very important ways that Web 2.0 paradigms and technologies can/should be brought to bear on inherently *business* problems.
One such problem is outlined fairly well HERE. Malcom Chisholm correctly suspects organizational "dark data" as not only the source of most information in the enterprise but also a the key untapped source for completing the intelligence picture.
According to Gartner,
Effective information management will be critical in the next decade, differentiating those enterprises that will implode under the infoglut from those that will use it to dominate the global economy. (Source: Gartner; June 2006; Spotlight on <?xml:namespace prefix = st1 /><st1:place w:st="on"><st1:City w:st="on">Enterprise</st1:City></st1:place> Information Management)
Contained in the catch-phrase "effective information management" are not just notions of version control and distribution automation. But there are also intricacies of extracting, finding and calculating relevant business intelligence from disparate and heterogeneous organizational information.
Chisholm describes the magical "dark integration" that have fancier terms on architecture diagrams but really end up signifying the same thing: human effort. People indexing data from scanned items into other systems. People scanning through a search result list to find what they *really* wanted. People performing audits or analysis or recommendations based on what they know, in their heads, to be what *matters*.
Because of the scope and pace of information generation, relying on the human element to perform the initial analysis is too little, too slow, too risky, too unpredictable, too inconsistent. What is needed is an Enterprise Information Management structure and approach that can leverage programmatic indexing, entity extraction, topical mapping, social inputs, sentiment and semantic awareness. These mashed up data sets become the new inputs in a new generation of composite business intelligence, real time decisioning, recommender, and suggestion engine technologies (to name a few).
The organization(s) that can bring these capabilities together in real business solutions will lead the next generation of adopting businesses to take over. The competitive advantage those organizations hold will be *that* significant.