Identity - Spheres of Influence

Jamie Lewis writes in Ends and Means: Identity in Two Worlds about the mis-understandings in the identity space. Trust, Identity and Policy are something I have given some thought to Creating trusted value chains. One reason for the confusion is that there are spheres of influence for identity. These spheres are in many ways different, I would characterize them as inside, outside and among. Inside is within a corporation and the notion of identity is controlled by the corporate directory(s). Outside are your customers coming into the enterprise and they have the individual identities that are not within the enterprise and also their identity is typically not attached to their enterprise. Among is when two organizations have a relationship the relationship may be managed by users with an identity within their organization but the basis of the relationship is among the two organizations.

The fundamental differentiation is the legal relationships, inside it is the employment contract we all have with our employer, outside is among us and the organization providing the service, and among it is the organizations involved.

Each of these spheres of legal control requires a different approach both technologically and socially. Unless we recognize them as different problems we will just have the circular conversations and talk at one another, rather than at the problem.

Mapping relationships - the next step in knowledge management?

Knowledge management has been described in many ways from the next big thing to an oxymoron. Much of the past work in knowledge management has focused on the information items rather than the relationships between the creators and consumers of the information. The first major application of the relationships among consumers and producers was Google. A simplistic explanation of the technology is that Google weights rankings by the relationships betweem people who produce and consume information.

There have been a flurry of efforts to visualize relationships among people to better understand how information flows, some even visualize Google using its web api Google Browser. There are several other efforts underway to visualize relationships by examining e-mail databases and looking at the to/from/cc relationships. One made it to C|Net E-mail patterns map corporate structure | CNET News.com another discussed by Ross Mayfield is an open source project Apache Agora by Stefano mazzocchi. I have been working on my own visualization of RSS links between blogs Blog Mapping which is making very slow progress.

Several companies (informal groups) are starting to appear around this space, RossMayfield's SocialText, Semaview looking at relationships using FOAF, and Groxis. I feel that this is a sign the next wave is coming and its foundation is in the move from unstructured to structured information. Blogs are one of the key platforms for moving from unstructured to structured information. They are providing a rich ecosystem to experiment with various structured data and metadata technologies. How long will it be before blogs become part of all the major platforms?

The key change that structured data/metadata and hence relationship mapping brings to knowledge management is that it makes it easier to connect the loosely coupled relationships between producers and consumers. I believe the value of information (and hence knowledge) has a very direct trust relationship between who produces it and who consumes it. Blogs magnify this trust effect - if a trusted node publishes a piece and it is then picked up by several other blogs and discussed the "knowledge trust metric" is increased. In a recent posting by Jon Udell Degrees of Freedom he quotes Sam Ruby "Its just data" - yes but the relationships the data has enhances its value (either making it more or less true).

For our understanding of each other and the world we need to consider the relationships between information as this is the context where we build understanding and trust between each other. Visualizing the flow of information in both time and space is necessary to achieve both understanding and trust. Without these missing elements we will not make much progress.

Anything that helps to add understanding and builds trust in our strife torn world cannot be a bad thing...

Knowledge Networks..

We frequently look for use cases when studying communities of knowledge workers. The largest and most successful (IMHO) knowledge network today is the open source community. One of the few major companies to tap into it in a constructive manner is IBM, the rest have different degrees of antipathy towards the community. The tools provided by sites such as Freshmeat provide the framework for very effective collaboration and information sharing.

The knowledge captured and shared by the open source community is probably one of the most dynamic and dense public bodies of knowledge available today. The concept of open source and blogs are very similar around the open sharing of ideas and allowing the network (i.e. empowered individuals) to select and propagate useful code or posts.

The open source community is probably the largest R&D laboratory in existence today. It is a classic example of simple tools for collaboration providing more value than all the complex tools combined have ever produced. While companies such as RedHat have tapped into this and created successful businesses they have only tapped into a small fraction of the value that the network has produced. Is there an organizing principle here where value can be created. By value we do not necessarily mean financial value but rather new applications that will benefit the community as a whole. Applications that move beyond the creation of Linux and GNU to the mass of users of computers.

Simple tools - Just the facts

Effective Social Networks on Ross Mayfields blog brings up the point of the complexity of groupware tools - I would add to this the whole field of knowledge management as being guilty of the same sin. There is a self perpetuating myth that there needs to be a complex set of software to manage large amounts of unstructured information. The myth appears to be that this information is hugely valuable and needs sophisticated tools to ensure you can squeeze the last drop of value out of it.

One of my professors taught his class a key rule in any engineering project - always do a rough estimate for any calculation/project to get an idea of the size before doing a detailed analysis. This serves several purposes - makes sure you understand the variables in the problem and makes sure you do not misplace any zeros in the detailed analysis. Doing the same in any groupware or knowledge management application is a similarly revealing exercise - take a representative sample of information and reduce it to the key facts - it quickly shows there is not much there and what is there does not benefit from complex tools.

So what is all this unstructured data - mainly attempts to create the evidence to prove the few facts that we are all trying to reach agreement on. Perhaps we should be looking for tools to assemble evidence in favor of the few key truths that we all hope exist.

Thanks to Jon...

Thanks to Jon Udell for pointing out my graph project. Unfortunately I do not think it was ready for the attention. I have some ideas on how to make it more useable and a lot more robust.

He does bring up some interesting points - similar issues have been discussed on Ross Mayfield's Social Networks. Where does technology help form communities and when does it get in the way. Jon's take on WebX is that it was getting in the way, of creating a community. Blogs are much easier but need some way of providing feedback more like threaded discussions.

To create social networks that can exerts political pressure it is necessary to bind them together to exert the collective will. This binding may be very transient but to impact any political process it must exist.

Interactive version with deep crawling

Working under the principle that it is better to release than not I put up another version of the blog map thing (need to think of a name). It has the two types of crawling 1) Shallow takes the node you select and finds all the rss/rdf feeds listed on the page - quick and painless. 2) Deep: gets all the links on the page and follows them down to see if they have a RSS feed that matches the link. If so they are a blog and so save. - this is slow.... Have patience I hope you will be rewarded with pretty pictures. There are still many bugs but I am having fun playing with it so I thought I would share the fun.Graph Toool