Social Software and Bottom-Up Knowledge Management

Yesterday I was at the steering committe of the research project KnoSoS, in which i-merge is participating. KnoSoS stands for "Knowledge Sharing over Social Software", and is a joined applied research project of Memori and the VUB university. It's a very exciting research project, because the consortium will develop throughout the course of the project a Knowledge Management platform that will be built around social software tools, social networking principles, tagging, matching, visualization,... The system they are building will be about bottom-up knowledge management. This implies that matching people will be as important as matching information. And doesn't this is a far better way to get in touch with the reality of an organisation? Isn't the biggest part of the available knowledge and knowhow in a company or in a network not written down in documents but all in the heads of the people that are part of this network? Therefore, wouldn't it be better to build a system that gets you in touch with relevant data, as well as with a relevant expert, of whom you can presume that he has an answer to your problem? The fascinating thing about this platform that the KnoSoS team is developing ...

... is that they glue information from three different spaces:

1. The identity space: This is the database of member profiles. Most dating and social networking sites allow users to browse through other users profiels, based on a set of criteria that every user can define.
2. The Dyadic space: Whenever two people have a one-to-one contact in a given network, and this contact can be recorded (e.g. through skype, msn, email), than this conversation can be seen as new "knowledge data" that is poured into the platform. This data has very usable metadata: when? between whom? what was the topic? what was the context of other conversations it formed part of (e.g a project)... all this metadata is implicitely availible. That's for instance why MS Outlook is the best knowledge management system we have: because the only thing we often remember about specific data that we're looking for, is the fact that a specific person send it to us around a given date.  A  knowledge management platform that is able to browse through all these dyadic interactions  has  a new source of knowledge at it's disposal that is not to be found in any traditional knowledge management system
3. The Group space: when a social networking site, offers the tools for their members to form ad hoc groups, (e.g. project teams)  than the value of this platform grows exponentially (that's known as Reed's Law). Whatever people do in an ad hoc group in any given network,.. they produce data and metadata:

• Data: e.g. a text they collaborated upon in a wiki, conversations they posted on a blog or through a mailinglist, events they inserted in a shared group agenda, etc...
• Metadata: e.g. when was this posted, who were the other team members, what's the version of this document, what's the project it formed part of? etc...

What i'm trying to get at is that the most exciting thing about this KnoSoS platform is that every interaction one has on a platform, whether in a dyadic form or through participation in a group, enriches the network identity of that person. That's the challenge: tying the idenity space with the other two spaces in such a way that my profile page can be the starting point to browse through all the interactions i had and vice versa: every interaction i had, should enable other users to click through to my profile page, on which they can click further on other interactions. In other words: identity becomes a dynamic concept: it grows over time and it consists of a growing quantity of threads and traces and links with data, people and groups. This is how data and persons get woven into an interactive network and this is the Knowledge Management challenge we're facing.