whoVIS editor-editor interaction visualization prototype, API for word provenance - Wiki-research-l

12 Mar 2015

Hi all, we produced a prototype of an editor-editor interaction network visualization for
individual articles, based on the word/tokens deleted and reintroduced by editors. It will
be presented as a demo at the WWW conference this year [1], but we would love to also get
some feedback on it from this list.  It's in an early stage and pretty slow when
loading up, so have patience when you try it out here:
http://km.aifb.kit.edu/sites/whovis/index.html, and be sure to read the "how to"
section on the site. Alternatively you can watch the (semi-professional) screencast I did
:P, it explains most of the functions.

The (disagreement) interactions are based on a extended version of the extraction of
authorship we do with wikiwho [2], and the graph drawing is done almost exactly after the
nice method proposed by Brandes et al. [3] . The code can be found at github, both for the
interaction-extraction extension of wikiwho [4] and the visualization itself [5], which
basically produces an json output for feeding the D3 visualization libraries we use.  We
have yet to generate output for more articles, so far we only show a handful for
demonstration purposes. The whole thing also fits nicely (and was supposed to go along)
with the IEG proposal that Pine had started on editor interaction [6] .

word provenance/authorship API prototype:

Also, we have worked a bit on our early prototype for an API for word
provenance/authorship:
You can get word/token-wise information from which revision what content originated (and
thereby which editor originally authored the word) at
http://193.175.238.123/wikiwho/wikiwho_api.py?revid=<REV_ID>"&am…
(<ARTICLENAME>  -> name of the article in ns:0, in the english wikipedia,
<REV_ID> -> rev_id of that article for which you want the authorship information,
format is currently only json)
Example:
http://193.175.238.123/wikiwho/wikiwho_api.py?revid=649876382&name=Laur…

Output format is currently:
{"tokens": [{"token": "<FIRST TOKEN IN THE WIKI MARKUP
TEXT>", "author_name": "<NAME OF AUTHOR OF THE TOKEN>",
"rev_id": "<REV_ID WHEN TOKEN WAS FIRST ADDED>"},
{"token": "<SECOND TOKEN IN THE WIKI MARKUP TEXT>",
"author_name": "<NAME OF AUTHOR OF THE TOKEN>",
"rev_id": "<REV_ID WHEN TOKEN WAS FIRST ADDED>"},
{"token": "<THIRD TOKEN …
… ], "message": null, "success": "true",
"revision": {"article": "<NAME OF REQUESTED ARTICLE>",
"time": "<TIMESTAMP OF REQUESTED REV_ID>", "reviid":
<REQUESTED REV_ID>, "author": "<AUTHOR OF REQUESTED
REV_ID>"}}

DISCLAIMER: there are problems with getting/processing the XML for larger articles right
now, so don't be surprised if that gives you an error sometimes (i.e. querying
"Barack Obama" for instance and similar sizes will *not* succeed for higher
revision numbers). Also, we are working on the speed and providing more precomputed
articles (right now almost all are computed on request, although we save intermediary
results). Still, for most articles it works fine and the output has been tested for
accuracy (cf. [2]).

At some point in the future, this API will also be able to deliver the interaction data
that the visualization is build on.

I'm looking forward to your feedback :)

Cheers,
Fabian

[1] http://f-squared.org/wikiwho/demo32.pdf
[2] http://f-squared.org/wikiwho/
[3] http://dl.acm.org/citation.cfm?id=1526808
[4] https://github.com/maribelacosta/wikiwho
[5] https://github.com/wikiwho/whovis
[6]
https://meta.wikimedia.org/wiki/Grants:IEG/Editor_Interaction_Data_Extracti…

--
Fabian Flöck
Research Associate
Computational Social Science department @GESIS
Unter Sachsenhausen 6-8, 50667 Cologne, Germany
Tel: + 49 (0) 221-47694-208
fabian.floeck@gesis.org<mailto:fabian.floeck@gesis.org>

www.gesis.org
www.facebook.com/gesis.org

[cid:E09117E0-16C9-4BCB-B1F9-D758CB4CE0D3]