Edit wars in Wikipedia

07/19/2011 ∙ by Róbert Sumi, et al. ∙ Budapest University of Technology and Economics 0

We present a new, efficient method for automatically detecting severe conflicts `edit wars' in Wikipedia and evaluate this method on six different language WPs. We discuss how the number of edits, reverts, the length of discussions, the burstiness of edits and reverts deviate in such pages from those following the general workflow, and argue that earlier work has significantly over-estimated the contentiousness of the Wikipedia editing process.



There are no comments yet.


page 4

This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.

I Introduction

The development of Wikipedia (WP) articles is not always a peaceful and collaborative process. This has long been recognized by the WP community, which calls extreme cases of disagreement over the contents of an article an edit war,111In the electronic version of this paper hyperlinks are direct references to WP. In the printed version, these are given in typewriter font. WP has developed specific guidelines for dealing with edit warring, such as the three revert rule, offers a variety of tags to warn about disputes, and even has a humorous listing of the lamest edit wars.

Perhaps the easiest way human readers can detect pages affected by warring (in the English WP, which is the one discussed unless explicitly stated otherwise) is to read through the discussion page (also known as the talk page) associated to each content page looking for telltale signs such as notices requesting cleanup, swearwords and name-calling. When the discussion grows heated, the length of the talk page can exceed the length of the article many times over, so that older discussions must be archived. Another way to detect controversy is to view the history of the page, which can show many war-like acts, in particular editors reverting the work of other editors.

Schneider et al [1] estimate that among highly edited or highly viewed articles (these notions are strongly correlated, see [2]) about 12% of discussions is devoted to reverts and vandalism, suggesting that the WP development process is highly contentious. In fact, once the great bulk of WP articles is considered, we find the editorial process far more peaceful: as we shall see, around 24k articles, i.e. less than 1% of the 3m articles available in the November 2009 English WP dump, can be called controversial. To sustain such far-reaching conclusions we can no longer rely on manual checking, so our primary interest is with the automatic detection of edit wars. Since our interest is with the entire WP process, of which the English WP is just the largest (and most mature) instance, we are primarily interested in language- and culture-independent methods that can be applied uniformly across the range of WPs. Therefore, our methods are based entirely on the history page, as opposed to the more human-readable talk page.

Previous works (including our own) aimed at the automatic detection of editorial conflict and edit wars is summarized in Section II. In Section III we discuss other indicators of controversiality and evaluate these in comparison to ours. We offer our conclusions in Section IV.

Ii Automatic conflict detection

Conflicts in WP were studied already both on the article and on the user level. Kittur [3] et al. computed article controversy from different page metrics (number of reverts, number of revisions etc.), Vuong et al. [4] counted the number of deleted words between users and used their “Mutual Reinforcement Principle” to measure how controversial a given article is. Both teams counted how many times dispute tags appeared in the history of an article, and used this as ground truth. While this is an excellent test in one direction (certainly recognition of controversiality by the participants is as good as the same recognition coming from an outsider), it is too narrow, as there can be quite significant wars that the community is unaware of or at least do not tag, as, e.g., in the articles on Gdańsk or Euthanasia. Note that by applying more lax criteria (i.e. not requiring the presence of overt conflict tags) our method will, if anything, overestimate the extent of controversy, strengthening our conclusion that there is much less conflict in WP than appears from sampling highly edited/viewed pages.

There are several papers which try to measure the negative links between WP editors in a given article and, based on this, attempt to classify editors into groups. The main idea of the method used by Kittur et al. 

[3, 5] is to count how many times an editor pair reverted each other. The more two editors reverted each other, the larger the conflict between them. As we shall see shortly, reverts are indeed central to the assessment of controversiality, but one needs to take into account not just the number of (presumably hostile) interactions, but also the seniority of the participants. Brandes et al. [6] assumed that users who do not agree with each other react very fast to edits by the others. The reciprocal value of the time elapsed between two consecutive edits increases the controversy between the two authors. In a more recent paper Brandes [7] counted the number of deleted words between editors and used this as a measure of controversy. West et. al. and also Adler et. al. have developed vandalism detection methods based on temporal patterns of edits [8, 9]. In both works the main assumption is that offensive edits are reverted much faster than normal edits, therefore by considering the time interval between an arbitrary edit and its subsequent reverts, one can classify vandalized versions with high precision.

Our own work (for a preliminary report, see [10]) was seeded by a manual sample of 40 articles, 20 selected for high controversiality, and 20 for low. Table I summarizes the number of reverts as detected in the text and in the comments222Each edit could be accompanied by a descriptive comment as Edit summary. (most reverts are detected by both methods).

Both txt Only Only Article
and cmt in txt in cmt titlee
4103 930 328 Global warming
2375 478 142 Homosexuality
1847 617 201 Abortion
1494 260 35 Benjamin Franklin
1425 437 130 Elvis Presley
1396 233 67 Nuclear power
1298 536 104 Nicolaus Copernicus
1071 211 51 Tiger
1036 248 58 Euthanasia
937 204 58 Alzheimer’s disease
870 192 50 Gun politics
836 172 23 Sherlock Holmes
689 213 49 Arab-Israeli conflict
659 496 138 Israel and the apartheid analogy
652 387 88 Liancourt Rocks
642 250 39 Schizophrenia
516 164 472 Gaza war
431 186 30 1948 Arab-Israeli war
416 73 9 Pumpkin
380 284 58 Gdańsk
318 158 20 SQL
162 24 10 Password
116 26 3 Henry Cavendish
109 29 4 Pension
81 29 4 Mexican drug war
74 37 10 Hungarians in Romania
70 14 4 Markov chain
70 12 1 Mentha
47 20 6 Foucault pendulum
40 5 6 Indian cobra
32 15 1 Harmonium
30 9 1 Infrared photography
29 4 1 Bohrium
24 34 5 Anyos Jedlik
11 6 2 Hungarian forint
10 3 1 Hendrik Lorentz
9 3 1 1980s oil glut
7 1 0 Deutsches Museum
4 0 0 Ara (genus)
0 0 0 Schlenk flask
TABLE I: Number of reverts detected. The upper part corresponds to a group of pages with severe conflicts (except those in italics); below the horizontal line there are peaceful pages (except those in italics).

Given the number and distribution of false positives and negatives (typeset in italics) it is clear from Table I that the raw revert statistics do not yield a clear cutoff-point we could use to distinguish controversial from non-controversial articles. Rather than building a complex but arbitrary formula that includes different factors that are expected to correlate with controversiality, our goal is to base the decision on very few parameters – ideally, just one.

Let be stages in the history of an article. If the text of revision coincides with the text of revision , we considered this a revert between the editor of revision and respectively. We are interested in disputes where editors have different opinions about the topic, and do not reach consensus easily.

Let us denote by the total number of edits in the given article of that user who edited the revision . We characterize reverts by pairs , where denotes the editor who makes the revert, and refers to the reverted editor (self-reverts are excluded). Fig.2 represents the revert map of the non-controversial Benjamin Franklin and the highly controversial Israel and the apartheid analogy articles. Each mark corresponds to one or more reverts. The coordinates of the marks are the total number of edits of the reverter () and the reverted editor (). Clearly, the disputed article contains more reverts between editors having large edit numbers than the uncontroversial article.

Fig. 1: Revert maps of the articles Benjamin Franklin (left) and Israel and the apartheid analogy (right). and are the total number of edits of the reverter and reverted editor respectively. The size of the mark is proportional to the number of reverts between them.
Fig. 2: Maps of mutual reverts in the same articles as in Fig. 2.
Fig. 1: Revert maps of the articles Benjamin Franklin (left) and Israel and the apartheid analogy (right). and are the total number of edits of the reverter and reverted editor respectively. The size of the mark is proportional to the number of reverts between them.

The revert maps already distinguish disputed and non-disputed articles, and we can improve the results by considering only the cases, in which two editors revert each other mutually, hereafter called mutual reverts. This causes little change in disputed articles (compare the right panels of Fig.2 to that of Fig.2), but has great impact on non-disputed articles (compare left panels).

Based on the rank (total edit number within an article) of editors, two main revert types can be distinguished: when one or both of the editors have few edits to their credit (these are typically reverts of vandalism since vandals do not get a chance to achieve a large edit number as they get banned by experienced users) and when both editors are experienced (created many edits). In order to express this distinction numerically, we use the lesser of the coordinates and , so that the total count includes vandalism-related reverts as well, but with a much smaller weight. Thus we define our raw measure of controversiality as

Once we developed our first autodetection algorithm based on , we iteratively refined the controversial and the noncontroversial seeds on multiple languages by manually checking pages scoring very high or very low. In this process, we improved in two ways: first, by multiplying with the number of editors who ever reverted mutually (the larger the armies, the larger the war) and define and second, by censuring the topmost mutually reverting editors (eliminating cases with conflicts between two persons only). Our final measure of controversiality is thus defined by


One conceptually easy (but in practice very labor-intensive) way to validate is by simply taking samples at different values and counting how many pages are found. As can be seen from Table II, in a binary manual classification between oncontroversial and ontroversial pages the number of controversial pages increases monotonically with increased from a low of 27% war to a high of 97% as goes from 50 to 31,000. (We note here that some of the truly controversial pages such as Anarchism have in excess of 10.)

50 22 8 44037 27 24
100 16 14 34112 47 20
180 15 15 26912 50 17
320 14 16 20763 53 13
560 14 16 15683 53 11
1000 12 18 11732 60 9
5600 1 29 4314 97 4
31000 1 29 1368 97 1
TABLE II: For a given level of , number of oncontroversial, ontroversial pages in random samples of 30, otal number of pages with greater , % of controversial pages, estimated total number of controversial pages with greater (numbers are in kilo and rounded to two significant digits).

We have checked this measure for six different languages (eventually leaving Romanian out for lack of data) and concluded that its overall performance is superior to other measures, including the presence of tags marking controversiality (see Section III).

Iii Other indicators of controversiality

There are many plausible candidates for measuring controversiality, such as the number of edits, the number of (mutual) reverts, the number of ‘controversial’ tags, and of course variants of our own measure . Table III compares the various methods for precision in the top 30 for each language (except Romanian for paucity of controversial pages), and for each method considered.

WP #e #r #mr TC
cs 14 18 26 25 27 27 28
en 27 29 29 26 28 30 28
hu 4 27 28 23 29 24 30
fa 24 28 26 29 29 25 28
es 23 26 29 27 28 28 29
%av 61 85 92 87 94 89 95
TABLE III: Precision of controversiality detection in the top 30 based on number of edits #e, reverts #r, mutual reverts #mr, raw , , article tag count TC, and .

It comes as no great surprise that the roughest measures such as the number of edits are rather poor classifiers and ‘controversial’ tag count (TC) are quite good in the English WP. Unfortunately, these measures fail to generalize from English and Spanish to smaller WPs, and given the cultural differences, there is no assurance that as the smaller WPs mature these measures will become increasingly applicable. As it is, measure given by Eq. 1 above reduces the precision errors of the hitherto best classifier by over 50% (10 errors compared to the best known method. article tag count, which had 24). Of our own methods, the final -based classifier improves upon the initial method (counting mutual reverts, 15 errors) by a full third.

The true value of a classifier of course depends not just on precision, but also on recall. This is much harder to measure, since the bottom 90% of the sample is uncontroversial by any measure, and it would take tens of thousands of manual judgments on random samples to obtain reliable recall figures. Also, the threshold for large WPs is much lower than one could surmise from inspecting the top 30 pages, for example in English is 1,620,378 for the 30th most controversial page. We selected

for cutoff, which yields a very high controversy population, but if we were truly intent on optimizing the threshold we would probably go down to about

to see as much vandalism as pure edit warring. As a shortcut, we therefore plotted the second best classifier against (see Fig. 3), and sampled from the quadrants where they make different predictions.

Fig. 3: Scatter plot of article Tag Count (TC) vs. . Color coding is proportional to the logarithm of the number of articles lying within each cell

On the whole, articles with low tag count but high appear to be quite controversial, even if the participants themselves fail to tag the article for controversy. The opposite situation, with low

and high TC, is found very rarely. Besides an inherently lower precision and recall, there are some mechanical reasons why counting controversial tags is not a perfect method to detect disputes. There are many dispute-related tags and one has to decide which tags to count on a per-language basis. Page  

[11] contains all disputed tags, but some may indicate more serious conflict than others (for example compare {{Unreferenced}} to {{Disputed}}).

The limitations of the earlier proposals such as [3] and [4] are evident if we check the results. For example, [4] concluded that in the Podcast article “a significant amount of dispute occurred between the two pairs of users: (a) user and user Jamadagni; and (b) user and user Yamamoto Ichiro”. A closer look at the article reveals that user edited the given article only once, and his edit was a vandalism, because he multiplied several time the text of the article, creating a revision which was 20 times larger than the previous one. Jamadagni simply reverted, generating this way a large number of deleted words between them. Real, recurrent disputes cannot produce this large amount of deleted words, therefore they remain hidden. (This is not an extreme example, user is another vandal, who edited the article only once.)

Iv Conclusion, future directions

We proposed a new way to measure how disputed a WP article is. We did this because existing models have drawbacks, and only a small fraction of WP articles were analyzed with them. We analyzed the whole WP for different language versions, and ranked articles according to their controversy level. Altogether, the proposed measure fares considerably better than earlier proposals both for precision and recall, though this fact would not be evident to the observer restricted to the top 30 articles of the English WP. For example, in Romanian even TC fails rather spectacularly. Based on the results obtained by our classifier, we conclude that in most cases, the process of development of the articles is considerably peaceful and the number of conflict cases have been overestimated in previous works, compared to our estimation of less than 1% of articles to be a candidate for a serious conflict. Besides being a robust language- and culture-independent classifier, our method also yields a numerical ranking, which agrees well with human judgment.

While our method does well in separating out edit wars from vandalism at the high end, much work remains to be done for lower . Researchers interested in a better controversiality measure may go back to the other indicators of controversiality and mix these into the measure: the largest limiting factor is the number of manually truthed examples one is willing to create for training and testing, but if resources are pooled across teams or the judgement task could be automated (e.g. by the Mechanical Turk) this limitation can be overcome.

Our future goals include detection of pure (non-war-like) vandalism, a task made all the more important by the high degree of vandalism we see. Another goal is the prediction of impending edit wars by monitoring the dynamics of – once a reasonable predictor is provided it will be possible to tag pages for impending conflict by robots.


This work was supported by the EU’s 7th Framework Programs FET-Open within ICTeCollective project no. 238597. Kornai also acknowledges support from OTKA grants #77476 (Algebra and algorithms) and #82333 (Semantic language technologies). Special thanks to Santo Fortunato for discussions and his help with data at early stages of this work.


  • [1] Schneider, Jodi and Passant, Alexandre and Breslin, John A Qualitative and Quantitative Analysis of How Wikipedia Talk Pages Are Used Proceedings of WebSci10 26-27 April, Raleigh, NC, 2010 http://journal.webscience.org/373
  • [2] J. Ratkiewicz, S. Fortunato, A. Flammini, F. Menczer and A. Vespignani Characterizing and modeling the dynamics of online popularity Physical Review Letters 105, article no. 158701, 2010.
  • [3] Aniket Kittur, Bongwon Suh, Bryan A. Pendleton, and Ed H. Chi. He says, she says: conflict and coordination in Wikipedia. In CHI ’07: Proceedings of the SIGCHI conference on Human factors in computing systems, pages 453–462, New York, NY, USA, 2007. ACM.
  • [4] Ba-Quy Vuong, Ee-Peng Lim, Aixin Sun, Minh-Tam Le, and Hady Wirawan Lauw. On ranking controversies in wikipedia: models and evaluation. In Proceedings of the international conference on Web search and web data mining, WSDM ’08, pages 171–182, New York, NY, USA, 2008. ACM.
  • [5] Bongwon Suh, E.H. Chi, B.A. Pendleton, and A. Kittur. Us vs. them: Understanding social dynamics in wikipedia with revert graph visualizations. Visual Analytics Science and Technology, 2007. VAST 2007. IEEE Symposium on, pages 163–170, 30 2007-Nov. 1 2007.
  • [6] Ulrik Brandes and Jürgen Lerner. Visual analysis of controversy in user-generated encyclopedias. Information Visualization, 7:34–48, March 2008.
  • [7] Ulrik Brandes, Patrick Kenis, Jürgen Lerner, and Denise van Raaij. Network analysis of collaboration structure in wikipedia. In Proceedings of the 18th international conference on World wide web, WWW ’09, pages 731–740, New York, NY, USA, 2009. ACM.
  • [8] Andrew G. West, Sampath Kannan, and Insup Lee. Detecting wikipedia vandalism via spatio-temporal analysis of revision metadata? In Proceedings of the Third European Workshop on System Security, EUROSEC ’10, pages 22–28, New York, NY, USA, 2010. ACM.
  • [9] B. Thomas Adler, Luca de Alfaro, and Ian Pye. Detecting Wikipedia Vandalism using WikiTrust: Lab Report for PAN at CLEF 2010 In Notebook Papers of CLEF 2010 LABs and Workshops, 22-23 September, Padoa, Italy, 2010. ISBN 978-88-904810-0-0.
  • [10] R. Sumi, T. Yasseri, A. Rung, A. Kornai, and J. Kertész. Characterization and prediction of Wikipedia edit wars In Proceedings of the ACM WebSci’11 June 14-17, Koblenz, Germany, 2011.
  • [11] http://en.wikipedia.org/wiki/Wikipedia:Template messages/Disputes