Understanding the Growth of Wikipedia via WikiProjects
Wikipedia is an evolving social machine that has gone through a series of phases similar to the development of the Web. We are now in a position where studying Wikipedia – whether this is to examine its successes, failures, current state of health, or future direction – requires understanding it as a social machine within an entire eco-system of interactions.
The metrics and methods that we use must take into consideration how Wikipedia is at the interface of multiple streams of input (and output), and consequently, we must analyse Wikipedia as a stream of interaction, rather than as a separate system that operates independently from other Web Systems.
Our thesis is that in the initial phase of Wikipedia, there was a substantial uptake of individuals able to contribute to a relatively small set of articles. Whilst the article count continues (along with the number of active Wikipedians), we start to observe knowledge saturation, which can often be misconstrued as a social machine health issue. Whilst the number of active editors is not growing at the rate of before, what we are observing is a growth in activity and interaction beyond the original purpose of the system. This model of Wikipedia growth is somewhat demonstrated by the visualisation of Wikipedia project growth; the community initially contribute to article content, then turns towards discussion activity.
Given these observations, we consider two strands of research that may yield worthwhile insights into understanding Wikipedia, as well as the space of social machines, in general. First, we will develop metrics that better represent the health and functionality of Wikipedia as a social machine . Second, we wish to study Wikipedia as a stream of activity across multiple social machines. We are already researching it’s role within the propagation of information from social networking platforms, but we wish to understand the impact to and from Wikipedia at the macro level (of the Web).
Below is our initial analysis of the Wikipedia Project ecosystem, we have extracted all edits made over the duration of a Wikipedia project, and compared it with the number of discussions and contributors. As the video shows, the growth of a project tends to follow a similar pattern; a project is first bootstrapped mostly by edits, by a small number of individuals. Overtime, more people contribute, and the number of discussions start to overtake the number of edits made.
• We want to consider Wikipedia as a stream of interaction, and similarly, we will consider the other social machines in this way.
• What can we say about the dynamics of Wikipedia if we look at it as a stream? Taking a systems-wide approach, can we study the diffusion of information across social machines, and what are the implications for Wikipedia?
• Can we identify the existence and workflows of procedural knowledge? Does the passage of information from external systems lead to the creation, addition, modification, or removal of knowledge within Wikipedia? Conversely, how is knowledge from Wikipedia diffused amongst other social machines?
• Are there patterns (or anti-patterns) that can be associated with different types of knowledge and/or topics?
• Ultimately, studying Wikipedia as a stream of information and interaction invokes a new paradigm, Re-assembling the Web to study it via its components