We've spent some time studying how companies can successfully use Web 2.0 to encourage participation among those both within and outside their walls. It's a way of achieving business goals, such as collaboratively developing content, solving technical problems, delivering marketing messages, and even predicting the outcome of uncertain events. We documented these findings in our article "Six ways to make Web 2.0 work."
Using Web 2.0, users explicitly participate in creating information. But users are also implicitly creating information, i.e., leaving behind a digital trail as they work on-line. Some have called this information "exhaust data." We're seeing an increasing number of applications that leverage implicit information, which was often explicitly created for other purposes. For example:
Google's PageRank system for ranking the relevance of search results relies on the links between Web pages. These links were originally created to allow people to move from one piece of Web content to another, but have been re-purposed to help determine the order in which to display search results.
Other systems can automatically extract the connections among people based on mining the implicit patterns in their communication, such as who is emailing whom. These systems contrast with "social networks," in which users must explicitly list connections with others in their social graph.
Systems can determine individuals' interests and expertise, based on the information that they write or post in electronic communications, e.g., in their blogs, or even email. Traditional corporate profiles, based on users explicitly entering their areas of expertise or interest, become quickly outdated. Systems based on current communications would be much more up-to-date.
What's interesting is that these examples suggest a corollary to our oft-quoted finding that "What's in the workflow is what gets used": Don't put extra tasks into someone's workflow. Whenever possible, leverage the implicit information generated by their workflow. It's easier to collect than explicit information.
While this makes sense, I think the example of social networks that you gave is heading in that direction as well. It is currently in the forming state (connections being made), but all webpages must go through that process of linking to others at some point. Once that linking process is complete, there is a wealth of ‘exhaust data’ generated from these websites as well (e.g., how often specific users sites are visited, trend tracking, etc)
Posted Jun 10, 07:22 PM by Birju Pandya