Review for Power of the Few vs. Wisdom of the Crowd: Wikipedia and the Rise of the Bourgeoisie

Summary: Describes a straight-forward, descriptive statistics analysis of the
distribution of work in the popular Wikipedia collaborative online system. The
main result, among many interesting results, indicates that initially so called
`elite' users contribute most changes to the content in the system, while later
their percentage of the total contributions declines. Interestingly, this
decline is not due to a decrease in activity by the elite, but rather by an
increase in contributions made by low-edit users (the so called `bourgeoisie').
This shift was studied across several different factors such as group type
(peer-selected vs. data-driven) and granularity of edit made (edit count vs.
word addition/deletion). The first paragraph of the discussion section on p. 7
provides an excellent, succinct summary of the findings. In order to help
determine if these results generalize to similar social collaborative online
systems, the authors conducted a similar analysis with del.icio.us, the popular
collaborative bookmark tagging site. They found a similar shift in
distribution of work even though del.icio.us and Wikipedia have some
fundamental differences in how users participate and collaborate. The authors
conclude by hypothesizing that this pattern may be ubiquitous in similar online
collaborative systems.

Contribution: The main contribution is the result that there seems to be a
distinct pattern in the evolution of online collaborative systems.

Review: The research methods and analyzes are sound. The authors studied the
distribution of work in the evolution of an online collaborative system from
multiple levels of granularity of both user groups and various edits. The
study, analysis, and prose are quite straight-forward and the authors interpret
the results well. While the interpretations of the results are interesting,
some have been shown to be insignificant, either by prior work (see missing
reference in the first paragraph of the subsection titled `Bots made
maintenance easier' on p. 3) or by these authors themselves (see sentence
starting `However, ....' at the top of the right column on p. 6). These
interpretations should be removed as they are not defendable and therefore
appear as noise. However, the reference (p. 3) and the results of additional
analysis (top, right column of p. 6) should remain. While the results are
interesting and well interpreted sociologically, the paper gives the practical
utility of these results and their implications for the design and sustenance
of online collaborative systems only a peripheral focus. The paper can be
improved by reducing the number of graphs (as the text is effective at making
the main results salient) and replacing them with a deeper discussion of the
design implications of the results. The authors might evolve the last
paragraph of the discussion section which briefly mentions these implications
into a new section. Lastly, while the related work section on p. 2 does tell a
story, it does not motivate the current study or put it in perspective with
respect to the related work.