Wikimania Social Machines Hack Event

A short summary and reflection of the Wikimania social machines hack event which happened on the 24/25th May 2014 in London.

The Wikimania Social Machines hack fest was a fantastic forum to initiate the discussion around the notion of social machines outside the core community of SOCIAM and Web Science researchers. Comprising of a two-day event, the first day focused around getting to know the attendees and their interests, which directly fed into a series of video conference calls with a number of key members of the Wikimedia organisation. Given the varying ‘Wikipedia’ knowledge of the attendees, the conference calls were an ideal opportunity to understand how Wikipedia operates, and get feedback from those outside the core community of editors and administrators.

Perhaps not so surprisingly, the topics discussed during the video calls revealed that Wikipedia is extremely complex, and even some of the core members of the community did not know all of the smaller processes, policies and idioms that embody and govern Wikipedia. The insights offered by Wikimedia's such as Philippe Beaudette, Aaron Halkfaker and Fabrice Florin told a story of a complex socio-technical ecosystem of social machines. similar to other environments we have been studying (take Twitter for instance), Wikipedia comprises of multiple social machines, each with their own human-machine processes, goals, motivations, and outcomes. From an observational perspective, understanding Wikipedia becomes an issue of granularity (this is something that really needs to be unpacked and understood), whilst it is possible to observe and analyse the system as a whole with metrics such as page edits and views, does this mask over the more subtle processes that are critical to the operation of Wikipedia?

The teleconferences and demonstrations by Ed Saperia and fellow Wikimedia representatives offered some really interesting views on how Wikipedia is being used far beyond the original design of an online knowledge base. Wiki-projects, badges, rating systems, Wikipedian profiles, Wikipedia tools, bots, talk, these are all emergent phenomenon that had grown from the initial idea of a collaborative environment where anyone with access to the Web could add and modify a Wiki page. The development of these socio-technical processes are the emergence of new social machines supported via the Wikipedia platform (which itself is a socio-technical system, supported by human and machine interactions). Take for example Wiki-projects, these are entire communities within the Wikipedia platform which emerge around a given topic or set of build around. We were shown the Medicine Wikipedia project, which comprises of a large number of medical articles in over 250 languages, produced and curated by an growing number of individuals, from medical students to practitioners. Within this project, which itself is a Wikipedia of medicine are mission statements and project goals which drive the progress of article quality and completeness. This progress is supported by the community, who engage with constant discussion using the ‘talk’ function, where new ‘users’ are greeted and introduced to the project.

However, ‘Wikiproject Medicine’ is not a unique case, there are hundreds if not thousands of emerging and active Wiki-projects. What makes this interesting is understanding the initial inertia for the formation of this project, how these projects attract new members, and how they actively sustain the commitment of the community. We also need to consider how these communities are developing their own technical and social mechanisms to support their work, and how these become adopted by other communities and projects over time. For instance, the medical project has developed a number of custom templates and navigation bars, which overtime may be adopted by other projects. There has also been consensus to form an organisation structure consisting of departments, participants and task forces, illustrating the formation and alignment of social processes.

Tying this back to the studies we've been doing in SOCIAM and citizen science analysis, a great cross-over is examining the relationship between talk and task (editing) (although as Ed pointed out, one user’s talk is another user’s task). Our analysis of citizen science projects have shown that like Wikipedia, online discussion between participants have become an emergent phenomenon, offering the community a means to chat, and in some cases, discover and confirm new scientific insights. What we also observe is the relationship of those talking, and those tasking, and typically, with a strong positive correlation between active talkers and taskers. Whilst Wikipedia is less prescribed in terms of an individual task and workflow, is it possible to observe the same talk-task characteristics identified in other online communities?

There is real potential here for SOCIAMers here, working with Wikipedia provides us the opportunity to actively observe the emergent phenomenon that led to the creation of social machines, and could prove to be an suitable environment for us to test out theories and mechanisms to support and potentially build new social machines.