On Tue, Jan 22, 2013 at 6:01 PM, Arjuna Rao Chavala
<arjunaraoc(a)gmail.com> wrote:
Both are key measures of activity for the outcome of
Wikipedia, that is
sharing the sum of human knowledge. If there are not many page views
resulting from less number of readers, there will be less enthusiasm on the
part of Editors to contribute. If there are more page views, more people
will be interested to become Editors. Given the nature of these metrics and
different ranges, each one will not be a reliable measure by itself as it
is the interaction in the Wikipedia eco-system that will be a more
appropriate measure. I also heard during some of wiki interactions that
when Chinese language wikipedia was banned in China, the number of editors
fell a lot.
When I realized that "Activity" is a product of the entities I
understood that the two entities are commutative. However, there was a
particular aspect which puzzled me - the number of views/viewers is a
function of the richness of the content. In other words, while that
value can certainly be influenced by the language community, it cannot
be controlled. Against that, the number of edits is a value that is
under the sphere of control of a language community. And, within that,
edits can perhaps be classified (in context of whether the data
available facilitates that deep dive) as : human and bots. Within the
human-edited subset, there are ways to visualize the trend of data
edits. Kiran did a bit of this way back -
<http://jace.zaiki.in/tag/mwclient>
TL;DR : the measure of activity could perhaps be accurately reflected
when specific data points around editing are considered rather than
using a relationship with views.
--
sankarshan mukhopadhyay
<https://twitter.com/#!/sankarshan>