Social Network Analysis of Wikipedia

List overview All Threads
Download

newer

older

Introduction and a simple question

The Wikimedia Research Newsletter...

Jeremy Foote

5 Sep 2012 5 Sep '12

9:43 p.m.

I am a brand new Master's student at Purdue. For my Social Network Analysis class, I'm thinking about doing a project about whether a Wikipedian's centrality in a network can be used as a predictor of future participation. I've spent the afternoon looking for relevant literature. I found the very interesting "Validity Issues in the Use of Social Network Analysis with Digital Trace Data" by Howison, Wiggins, and Crowston and "Network analysis of collaboration structure in Wikipedia" by Brandes et al. I'm wondering if there are other papers about how to translate Wikipedia into a network structure, or even more specifically relating to node-level centrality measures and participation measures. Very many thanks, Jeremy Foote

Attachments:

attachment.htm (text/html — 1012 bytes)

Show replies by date

Joe Corneli

5 Sep 5 Sep

10 p.m.

Hi Jeremy: This seems to be implied in your approach... but to reiterate: since the encyclopedia is a graph, it would be most useful to start by looking at centrality measures etc. there. If you define the "social" network (of people, where A<->B if they've edited the same page), then clearly the two graphs are "related". But what might you learn by comparing the two? - You might find people who have contributed a lot to a lot of articles (but especially more central ones?) - You might find people who have contributed a lot to only (relatively) few articles (perhaps especially more fringe topics?) - You might find people who have contributed a little to a lot of articles (how does this relate to centrality?) - and of course you might find lots and lots of people who have contributed a little to just a few articles (how does this relate to centrality?) That would likely be good enough for a first paper... Thinking further, each of the cells in this little 2x2 grid is presumably an "attractor", so you don't necessarily expect to see people cross from one cell to another. But it could be interesting to for these crossings... and here, for a second research phase, you could try using the "social" data to see how particular things like conversations on talk pages or reverts can motivate or demotivate people... I think there's already a lot of work on this, but I'm not sure if it's backed up by the same sort of data intensive approach you seem to be proposing. Good luck, and I'll be interested to know more about your work and results as this develops. Joe On Wed, Sep 5, 2012 at 8:43 PM, Jeremy Foote <foote0(a)purdue.edu> wrote:

...

Brian Keegan

10:05 p.m.

...

-- Brian C. Keegan Ph.D. Student - Media, Technology, & Society School of Communication, Northwestern University Science of Networks in Communities, Laboratory for Collaborative Technology

Dario Taraborelli

10:09 p.m.

you should also check out: Laniado, David, Riccardo Tasso, Y. Volkovich, and Andreas Kaltenbrunner. When the Wikipedians talk: network and tree structure of Wikipedia discussion pages. In Proceedings of the Fifth International AAAI Conference on Weblogs and Social Media (ICWSM '11), 177-184, 2011. http://www.aaai.org/ocs/index.php/ICWSM/ICWSM11/paper/viewPDFInterstitial/2… summarized here: https://meta.wikimedia.org/wiki/Research:Newsletter/2011-07-25#The_anatomy_… Dario On Sep 5, 2012, at 1:05 PM, Brian Keegan wrote:

...

There's a good amount of research Jullien 2012 has an excellent (although by no means exhaustive) lit review of extant Wikipedia research including many network analysis papers: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2053597 Welser, et al. 2011 use network analysis approaches to identify and differentiate users social roles: http://www.connectedaction.net/wp-content/uploads/2011/04/Welser.Cosley.plu… Antin, et al. 2012 use some centrality-like metrics to measure the diversity of editing behavior: http://faculty.poly.edu/~onov/Antin_Chehsire_Nov_WPP_CSCW_2012.pdf Kane 2009 on how network position influences article quality: http://www.profkane.com/uploads/7/9/1/3/79137/kane_2009_ocisa.pdf Kane, et al. 2012 on how membership turnover/retention influences article quality: http://www.samransbotham.com/sites/default/files/RansbothamKane_WikiDemotio… <shameless self promotion> Descriptive analysis of Wikipedia's response and networks to the 2011 Tohoku earthquake and tsunami: http://www.brianckeegan.com/papers/WikiSym11.pdf Developing a statistical model of whether Wikipedia collaborations as a bipartite network of editors and authors are more strongly influenced by features of editors or features of articles: http://www.brianckeegan.com/papers/CSCW12.pdf Developing a unipartite network of Wikipedia collaborations as "document passing" network among editors on a single article: http://www.brianckeegan.com/papers/WikiSym12.pdf On Wed, Sep 5, 2012 at 7:43 PM, Jeremy Foote <foote0(a)purdue.edu> wrote: I am a brand new Master's student at Purdue. For my Social Network Analysis class, I'm thinking about doing a project about whether a Wikipedian's centrality in a network can be used as a predictor of future participation. I've spent the afternoon looking for relevant literature. I found the very interesting "Validity Issues in the Use of Social Network Analysis with Digital Trace Data" by Howison, Wiggins, and Crowston and "Network analysis of collaboration structure in Wikipedia" by Brandes et al. I'm wondering if there are other papers about how to translate Wikipedia into a network structure, or even more specifically relating to node-level centrality measures and participation measures. Very many thanks, Jeremy Foote _______________________________________________ Wiki-research-l mailing list Wiki-research-l(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l -- Brian C. Keegan Ph.D. Student - Media, Technology, & Society School of Communication, Northwestern University Science of Networks in Communities, Laboratory for Collaborative Technology _______________________________________________ Wiki-research-l mailing list Wiki-research-l(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l

Philippe Beaudette

11:30 p.m.

I would also point you toward the (tangentially) related work by Dr. James Fowler at UCSD - he's done similar work on predictive analysis on Facebook. http://jhfowler.ucsd.edu/ pb ___________________ Philippe Beaudette Director, Community Advocacy Wikimedia Foundation, Inc. 415-839-6885, x 6643 philippe(a)wikimedia.org On Wed, Sep 5, 2012 at 1:09 PM, Dario Taraborelli < dtaraborelli(a)wikimedia.org> wrote:

...

There's a good amount of research Jullien 2012 has an excellent (although by no means exhaustive) lit

review of extant Wikipedia research including many network analysis papers:

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2053597 Welser, et al. 2011 use network analysis approaches to identify and

differentiate users social roles:

http://www.connectedaction.net/wp-content/uploads/2011/04/Welser.Cosley.plu…

> Antin, et al. 2012 use some centrality-like metrics to measure the diversity of editing behavior: > http://faculty.poly.edu/~onov/Antin_Chehsire_Nov_WPP_CSCW_2012.pdf

> Kane 2009 on how network position influences article quality: > http://www.profkane.com/uploads/7/9/1/3/79137/kane_2009_ocisa.pdf

> Kane, et al. 2012 on how membership turnover/retention influences article quality:

http://www.samransbotham.com/sites/default/files/RansbothamKane_WikiDemotio…

> <shameless self promotion> > Descriptive analysis of Wikipedia's response and networks to the 2011 Tohoku earthquake and tsunami: > http://www.brianckeegan.com/papers/WikiSym11.pdf

> Developing a statistical model of whether Wikipedia collaborations as a bipartite network of editors and authors are more strongly influenced by features of editors or features of articles: > http://www.brianckeegan.com/papers/CSCW12.pdf

> Developing a unipartite network of Wikipedia collaborations as "document passing" network among editors on a single article: > http://www.brianckeegan.com/papers/WikiSym12.pdf

> On Wed, Sep 5, 2012 at 7:43 PM, Jeremy Foote <foote0(a)purdue.edu> wrote: > I am a brand new Master's student at Purdue. For my Social Network Analysis class, I'm thinking about doing a project about whether a Wikipedian's centrality in a network can be used as a predictor of future participation. I've spent the afternoon looking for relevant literature. I found the very interesting

> "Validity Issues in the Use of Social Network Analysis with Digital Trace Data" by Howison, Wiggins, and Crowston

and "Network analysis of collaboration structure in Wikipedia" by Brandes et

al.

> I'm wondering if there are other papers about how to translate Wikipedia into a network structure, or even more specifically relating to node-level centrality measures and participation measures.

> Very many thanks, > Jeremy Foote

_______________________________________________ Wiki-research-l mailing list Wiki-research-l(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l

> -- > Brian C. Keegan > Ph.D. Student - Media, Technology, & Society > School of Communication, Northwestern University

> Science of Networks in Communities, Laboratory for Collaborative Technology

_______________________________________________ Wiki-research-l mailing list Wiki-research-l(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l

Chitu Okoli

11:33 p.m.

...

_______________________________________________ Wiki-research-l mailing list Wiki-research-l(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l

Jeremy Foote

6 Sep 6 Sep

3:37 a.m.

Very, very many thanks to all! On Wed, Sep 5, 2012 at 5:33 PM, Chitu Okoli <Chitu.Okoli(a)concordia.ca>wrote;wrote:

...

Searching the WikiLit site for "social network" (in quotes) also brings up many relevant studies (though there are a few false matches): http://wikilit.referata.com/w/index.php?title=Special%3ASearch&search=%… ~ Chitu -------- Message original -------- Sujet: Re: [Wiki-research-l] Social Network Analysis of Wikipedia De : Dario Taraborelli ** Pour : Research into Wikimedia content and communities ** Date : 5 Septembre 2012 16:09:34 **** you should also check out: Laniado, David, Riccardo Tasso, Y. Volkovich, and Andreas Kaltenbrunner. When the Wikipedians talk: network and tree structure of Wikipedia discussion pages. In Proceedings of the Fifth International AAAI Conference on Weblogs and Social Media (ICWSM '11), 177-184, 2011. http://www.aaai.org/ocs/index.php/ICWSM/ICWSM11/paper/viewPDFInterstitial/2… summarized here: https://meta.wikimedia.org/wiki/Research:Newsletter/2011-07-25#The_anatomy_… Dario On Sep 5, 2012, at 1:05 PM, Brian Keegan wrote: There's a good amount of research Jullien 2012 has an excellent (although by no means exhaustive) lit review of extant Wikipedia research including many network analysis papers:http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2053597 Welser, et al. 2011 use network analysis approaches to identify and differentiate users social roles:http://www.connectedaction.net/wp-content/uploads/2011/04/Welser.Cosl… Antin, et al. 2012 use some centrality-like metrics to measure the diversity of editing behavior:http://faculty.poly.edu/~onov/Antin_Chehsire_Nov_WPP_CSCW_2012.pdf Kane 2009 on how network position influences article quality:http://www.profkane.com/uploads/7/9/1/3/79137/kane_2009_ocisa.pdf Kane, et al. 2012 on how membership turnover/retention influences article quality:http://www.samransbotham.com/sites/default/files/RansbothamKane_Wik… <shameless self promotion> Descriptive analysis of Wikipedia's response and networks to the 2011 Tohoku earthquake and tsunami:http://www.brianckeegan.com/papers/WikiSym11.pdf Developing a statistical model of whether Wikipedia collaborations as a bipartite network of editors and authors are more strongly influenced by features of editors or features of articles:http://www.brianckeegan.com/papers/CSCW12.pdf Developing a unipartite network of Wikipedia collaborations as "document passing" network among editors on a single article:http://www.brianckeegan.com/papers/WikiSym12.pdf On Wed, Sep 5, 2012 at 7:43 PM, Jeremy Foote <foote0(a)purdue.edu> <foote0(a)purdue.edu> wrote: I am a brand new Master's student at Purdue. For my Social Network Analysis class, I'm thinking about doing a project about whether a Wikipedian's centrality in a network can be used as a predictor of future participation. I've spent the afternoon looking for relevant literature. I found the very interesting "Validity Issues in the Use of Social Network Analysis with Digital Trace Data" by Howison, Wiggins, and Crowston and "Network analysis of collaboration structure in Wikipedia" by Brandes et al. I'm wondering if there are other papers about how to translate Wikipedia into a network structure, or even more specifically relating to node-level centrality measures and participation measures. Very many thanks, Jeremy Foote _______________________________________________ Wiki-research-l mailing listWiki-research-l@lists.wikimedia.orghttps://lists.wikimedia.org/mailman/listinfo/wiki-research-l -- Brian C. Keegan Ph.D. Student - Media, Technology, & Society School of Communication, Northwestern University Science of Networks in Communities, Laboratory for Collaborative Technology _______________________________________________ Wiki-research-l mailing listWiki-research-l@lists.wikimedia.orghttps://lists.wikimedia.org/mailman/listinfo/wiki-research-l _______________________________________________ Wiki-research-l mailing listWiki-research-l@lists.wikimedia.orghttps://lists.wikimedia.org/mailman/listinfo/wiki-research-l _______________________________________________ Wiki-research-l mailing list Wiki-research-l(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l

Tilman Bayer

5 Sep 5 Sep

10:08 p.m.

Hi Jeremy, there are quite a few papers which have done social network analysis of (mostly the English) Wikipedia; e.g. in the "Wikimedia Research Newsletter" we covered two which looked at centrality in different contexts: https://meta.wikimedia.org/wiki/Research:Newsletter/2012-06-25#Briefly ("'Central' users produce higher quality") https://meta.wikimedia.org/wiki/Research:Newsletter/2011-09-26#How_social_t… ("Closeness, PageRank, and eigenvector centrality were found to have the largest regression coefficients in predicting the outcome of an RfA") On Wed, Sep 5, 2012 at 12:43 PM, Jeremy Foote <foote0(a)purdue.edu> wrote:

...

-- Tilman Bayer Senior Operations Analyst (Movement Communications) Wikimedia Foundation IRC (Freenode): HaeB

4243

days inactive

4244

days old

wiki-research-l@lists.wikimedia.org

Manage subscription

7 comments

7 participants

tags (0)

participants (7)

Brian Keegan
Chitu Okoli
Dario Taraborelli
Jeremy Foote
Joe Corneli
Philippe Beaudette
Tilman Bayer