Dave Pollard puts together a proposal for what business weblogs should do. He makes the case for ease of use and describes how many of the typical knowledge mangement functions translate into the social software realm. This is something to bookmark and think about again.
A high-level spec for business weblogs and social software
... the key technical elements of Social Networking Enablement (SNE) are business weblogs (the repositories of personal knowledge) and social software (the tools that connect people and mine their knowledge). Following is a high-level specification for commercial development of such software. In organizations with structured work processes (manufacturers, banks etc.) these elements would supplement centralized, filtered knowledge repositories of best practices, policies and methodologies etc. In organizations with primarily unstructured work processes (consultants, engineers etc.) these elements could largely supplant centralized, filtered knowledge repositories and the tools that access them.
Corante Social Software was particularly interested in the fourth proposition for a Knowledge Traffic Management tool:
This tool would identify areas of knowledge sharing 'congestion' (people who are receiving an unmanageable number of requests for information, or not responding to requests on a timely basis), topics that are suddenly 'hot', and the adequacy of the enterprise's knowledge about those topics, people who are excessively isolated from others (few connections or exchanges), de facto experts and thought leaders who should be recognized (or, if they are outside the enterprise, perhaps hired), etc.
Similar capabilities have been mentioned in association with communities of practice. Rather than having users discover that someone (an expert, usually) is busy because they don't respond, have the system more fully integrated into the infrastructure, so that it "knows" when people are likely to respond. Essentially, he is suggesting that the software have built in social network analysis capabilities to identify experts (tool #1) and people who are disconnected from the network. This goes further still in suggesting that the system should also watch for increased traffic on a given topic - particularly if it is new to the taxonomy of the organization - to highlight new knowledge/ideas that might be of value to the company.
Is this business intelligence for "unstructured data?"