Posted on | November 13, 2007 | No Comments
I found a rather insightful commentary discussing some of Wikipedia’s fundamental problems in the Slashdot submission queue today. The commentary is presented more or less as an introspective series of rhetorical questions from the perspective of a disgruntled user.
I don’t find Wikipedia’s problems to be very interesting. While I hope that Wikipedia retains its strong editorial base, I’m mostly fascinated by the consistency of human nature and appreciate the ability to study it ‘in action’.
A few years ago, scientists began producing methods to model and map social networks. Social networks are of particular interest to many researchers because such networks consistently demonstrate themselves as being small world. Most researching were curious to see if phenomenon such as fame could be predicted, I just wanted to watch Sociology 2.0 take shape.
Small world networks can be mapped, this helps you to determine their efficiency, or how quickly instructions or information is shared amongst the nodes in a network. How quickly this happens depends entirely on the number of hops between nodes, or degrees of separation.
If the network is comprised of human beings, ‘gossip’ would be a perfect test of network efficiency. This is also commonly referred to as ‘six degrees of separation’, meaning, you know everyone else in the network through six recursive acquaintances. If you know six people, you know almost everyone else in the world by ‘proxy’.
When others were studying Myspace, the ‘blogosphere’ or other similar networks, I went to work on dissecting Wikipedia. Believe it or not, Wikipedia is one of the largest clusters of social networks on the Internet, a rather incidental result of massive collaboration. Wikipedia is fun to study beyond the content of its articles, its ever changing structure and thousands of small social networks are also of interest. Additionally, you can’t get monthly database dumps free from Myspace.
Because my work and research activities often involve global collaboration that does not benefit from the non-verbal parts of communication, I’m eager to develop methods that allow large scale free/open source communities achieve a reliable ‘bird’s eye’ view of their social health. Thus far, for the purposes of easy analysis, I categorize the nodes in a social network (I suppose I could just call them people) into three categories: reliable, available and masonic. Masonic in no way refers to the Free Masons, I’m alluding to the mortar used in masonry. I’ll describe these types below:
A reliable node is one that is not likely to change, often reliable nodes are leaders. They are consistent in endeavor and demonstrate frequent initiative.
Available nodes follow reliable nodes. They are dynamic in nature and inconsistent in endeavor. Available nodes subscribe to the initiative of reliable nodes and become the means to transform ideas into results.
Masonic nodes are similar to available nodes with the exception of demonstrating initiative. Masonic nodes become reliable-like with the exception of demonstrating consistency in endeavor. A masonic node is usually found functioning as a bridge between two networks, or creating a clone of an existing network with a slightly different agenda.
Of most interest to me is studying patterns that expose masonic nodes, which are invaluable to monitor the ‘social health’ of any given network. This involves watching where available nodes come and go as their endeavors change. That leads me to another curious phenomenon in networks, trust.
Trust cripples network efficiency, in any kind of network. If we think of computers, we talk of scenarios where machines are programmed to be intentionally rude and distrusting for security purposes. This is a condition that machines inherited from their creators, humans. Just like in a computer network, trust (or lack thereof) can break network efficiency by rendering any given node useless, a condition that is typically suffered by available nodes. If communication to or from the network breaks, an available node will join another network or become a masonic node, depending on its degree of initiative.
‘Break’ simply implies a lack of replies from the network, an inability to voice something or consistent misunderstandings when communicating with a network.
Patterns that evidence trust breaking are of interest, as of course such an event results in re-structuring of a network’s information flow.
Will I ever come up with some kind of collaboration system that incorporates a social heartbeat monitor? I doubt it. I am attracted to this type of research because it would help to solve many problems, such as schools being able to really understand how networks like Myspace effect the social structure of their student body, or allow collaborative efforts to be more productive by illustrating where things break (and possibly predicting such breaks).
The problem remains that the research itself is not static in methodology. To study Wikipedia, you need only to download database dumps or scrape edit and talk pages. To study similar concepts in networks like Myspace, you’d have to come up with different methods of getting and relating data. No single program could ‘do it all’. What remains are the methods, which thankfully (broadly) fall into the realm of prior art.
I’m thinking of producing some generic building blocks to help study ‘Sociology 2.0′, but still just kicking the ideas around. Some software exists that is helpful, such as W3C’s http engine , but most Sociologists are not C programmers. I’m also not overly fond of using software released under a license that proposes a creative work to be intellectual property.
Sure, I guess you could spend time mapping link relations and predict fame, but what fun is that? I really enjoy watching people be people. So, at least for me, the ‘clique’ clashes that evidence themselves on Wikipedia serve some useful purpose.
This is yet one more time consuming hobby that will probably never produce a dime, but it keeps me entertained.
Incidentally, while living in South East Asia, I’ve been studying whatever presents itself as interesting. The same social behaviors that I’ve described in this post are also consistently demonstrated by cockroaches.