to add some context to the present approach, you may remember that when we defined Editor Model metrics we started from the highest possible level of aggregation (i.e. all namespaces combined, archive table included). See rationale below from a previous email exchange:
we tried to stick to two general principles: 1) we want to count users making contributions to a project as a whole. Establishing that only “content activity” should be considered means that someone uploading a picture, editing a template, drafting an article outside of ns0 (we have a new Draft namespace), writing or contributing to a new policy, helping coordinate a wikiproject, i.e. all fundamental activities that contribute to the growth of the project, would be discounted as an editor. By this token, someone writing an entire article outside of the main namespace would not be included as an editor while a vandal fighter only reverting edits at the push of a button would be considered as a contributor. The point I’m trying to make is that establishing what “content” means is very arbitrary and we should have a measure of overall participation to a project, followed by more granular metrics by type of contribution (see next point). 2) instead of starting with a list of exclusions (i.e. we will only measure a subset of ns0 edits on articles meeting specific criteria such as countable pages), we will introduce breakdowns that inform us about specific types of activity. “Namespace” is a possible proxy for types of content, but not necessarily the best or the only one. One day, I’d like to be able to monitor active typo-fixers or template-editors, but I believe we should start from the highest possible level and count total activity or total unique editors before breaking them down. Adding a NS dimension or other criteria to filter top-level metrics sounds like a totally legitimate request as a metric breakdown. Dario > On Nov 4, 2014, at 12:55 PM, James Forrester <jforres...@wikimedia.org> wrote: > > Thanks Toby! :-) > > On 4 November 2014 12:38, Toby Negrin <tneg...@wikimedia.org > <mailto:tneg...@wikimedia.org>> wrote: > Created tracking bug -- please add yourselves to the cc if desired. > > https://bugzilla.wikimedia.org/show_bug.cgi?id=72973 > <https://bugzilla.wikimedia.org/show_bug.cgi?id=72973> > > -Toby > > On Tue, Nov 4, 2014 at 12:07 PM, James Forrester <jforres...@wikimedia.org > <mailto:jforres...@wikimedia.org>> wrote: > On 4 November 2014 12:00, Aaron Halfaker <ahalfa...@wikimedia.org > <mailto:ahalfa...@wikimedia.org>> wrote: > Understood for page creations. The metric is named "Page creations". We > ought to have a metric called "Content page creations" or "Unique content > page creators". > > Yeah, having both would be great but I don't want to demand the world on a > stick. ;-) > > One bit of complication: How do you feel about the draft namespace for > enwiki? Should it be included in content page creations? > > That should be in $wgContentNamespaces but unfortunately isn't (see the > config file > <http://noc.wikimedia.org/conf/highlight.php?file=InitialiseSettings.php>). > I'll get that fixed. > > As for edits, the correlation is so strong between edits to content and edits > to other namespaces that it doesn't matter which we use when looking for > trends.[1] > > 1. > http://meta.wikimedia.org/wiki/Research:Refining_the_definition_of_monthly_active_editors > > <http://meta.wikimedia.org/wiki/Research:Refining_the_definition_of_monthly_active_editors> > > Fair point. > > J. > -- > James D. Forrester > Product Manager, Editing > Wikimedia Foundation, Inc. > > jforres...@wikimedia.org <mailto:jforres...@wikimedia.org> | @jdforrester > > _______________________________________________ > Analytics mailing list > Analytics@lists.wikimedia.org <mailto:Analytics@lists.wikimedia.org> > https://lists.wikimedia.org/mailman/listinfo/analytics > <https://lists.wikimedia.org/mailman/listinfo/analytics> > > > > _______________________________________________ > Analytics mailing list > Analytics@lists.wikimedia.org <mailto:Analytics@lists.wikimedia.org> > https://lists.wikimedia.org/mailman/listinfo/analytics > <https://lists.wikimedia.org/mailman/listinfo/analytics> > > > > > -- > James D. Forrester > Product Manager, Editing > Wikimedia Foundation, Inc. > > jforres...@wikimedia.org <mailto:jforres...@wikimedia.org> | @jdforrester > _______________________________________________ > Analytics mailing list > Analytics@lists.wikimedia.org > https://lists.wikimedia.org/mailman/listinfo/analytics
_______________________________________________ Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
