Creating, destroying, and restoring value in Wikipedia

From WikiLit
Jump to: navigation, search
Publication (help)
Creating, destroying, and restoring value in Wikipedia
Authors: Reid Priedhorsky, Jilin Chen, Shyong Tony K. Lam, Katherine Panciera, Loren Terveen, John Riedl [edit item]
Citation: GROUP '07 Proceedings of the 2007 international ACM conference on Supporting group work  : 259-268. 2007 November 4-7. Sanibel Island, FL, United states. Association for Computing Machinery.
Publication type: Conference paper
Peer-reviewed: Yes
Database(s):
DOI: 10.1145/1316624.1316663.
Google Scholar cites: Citations
Link(s): Paper link
Added by Wikilit team: Yes
Search
Article: Google Scholar BASE PubMed
Other scholarly wikis: AcaWiki Brede Wiki WikiPapers
Web search: Bing Google Yahoo!Google PDF
Other:
Services
Format: BibTeX
Creating, destroying, and restoring value in Wikipedia is a publication by Reid Priedhorsky, Jilin Chen, Shyong Tony K. Lam, Katherine Panciera, Loren Terveen, John Riedl.


[edit] Abstract

Wikipedia's brilliance and curse is that any user can edit any of the encyclopedia entries. We introduce the notion of the impact of an edit, measured by the number of times the edited version is viewed. Using several datasets, including recent logs of all article views, we show that an overwhelming majority of the viewed words were written by frequent editors and that this majority is increasing. Similarly, using the same impact measure, we show that the probability of a typical article view being damaged is small but increasing, and we present empirically grounded classes of damage. Finally, we make policy recommendations for Wikipedia and other wikis in light of these findings.

[edit] Research questions

"We pose three specific research questions: 1. Creating value: Who contributes Wikipedia’s value? Is it the handful of people who edit thousands of times, or is it the thousands of people who edit a handful of times? 2. Impact of damage: What is the impact of damage such as nonsensical, offensive, or false content? How quickly is it repaired, and how much of it persists long enough to confuse, offend, or mislead readers? 3. Types of damage: What types of damage occur, and how often?"

Research details

Topics: Other content topics, Vandalism, Participation trends, Collaboration software [edit item]
Domains: Information systems [edit item]
Theory type: Analysis [edit item]
Wikipedia coverage: Main topic [edit item]
Theories: "Undetermined" [edit item]
Research design: Statistical analysis [edit item]
Data source: Computer usage logs, Wikipedia pages [edit item]
Collected data time dimension: Longitudinal [edit item]
Unit of analysis: Article view, Edit [edit item]
Wikipedia data extraction: Dump [edit item]
Wikipedia page type: Article [edit item]
Wikipedia language: English [edit item]

[edit] Conclusion

"Our view-based metrics let us both sharpen previous results and go beyond them. Others have shown that 1% of Wikipedia editors contributed about half of edits [6]. We show that 1/10th of 1% of editors contributed nearly half of the value, measured by words read. Others have shown that one type of damage was repaired quickly [20]. We show this for all types of damage. We also show what this result means for readers: 42% of damage is repaired almost immediately, i.e., before it can confuse, offend, or mislead anyone. Nonetheless, there are still hundreds of millions of damaged views. We categorize the types of damage that occured, show how often they occured, describe their potential impact on readers, and discuss how hard (or easy) they are to detect automatically. We give examples of especially impactful damage to illustrate these points. Finally, we show that the probability of encountering damage increased exponentially from January 2003 to June 2006. What are the implications of our results? First, because a very small proportion of Wikipedia editors account for most of its value, it is important to keep them happy, for example by ensuring that they gain appropriate visibility and status. However, turnover is inevitable in any online community. Wikipedia should also develop policies, tools, and user interfaces to bring in newcomers, teach them community norms, and help them become effective editors. Second, we speculate that the exponential increase in the probability of encountering damage was stopped by the widespread use of anti-vandalism bots. It is likely that vandals will continue working to defeat the bots, leading to an arms race. Thus, continued work on automatic detection of damage is important. Our results suggest types of damage to focus on; the good news is that the results show little subtlety among most vandals. We also generally believe in augmentation, not automation. That is, we prefer intelligent task routing [7] approaches, where automation directs humans to potential damage incidents, but humans make the final decision."

[edit] Comments

""We show that 1/10th of 1% of editors contributed nearly half of the value, measured by words read... [and] 42% of damage is repaired almost immediately, i.e., before it can confuse, offend, or mislead anyone." "[B]ecause a very small proportion of Wikipedia editors account for most of its value, it is important to keep them happy, for example by ensuring that they gain appropriate visibility and status. However, turnover is inevitable in any online community. Wikipedia should also develop policies, tools, and user interfaces to bring in newcomers, teach them community norms, and help them become effective editors." p.267-268 Wikipedia pages; request logs"


Further notes[edit]

Facts about "Creating, destroying, and restoring value in Wikipedia"RDF feed
AbstractWikipedia's brilliance and curse is that aWikipedia's brilliance and curse is that any user can edit any of the encyclopedia entries. We introduce the notion of the impact of an edit, measured by the number of times the edited version is viewed. Using several datasets, including recent logs of all article views, we show that an overwhelming majority of the viewed words were written by frequent editors and that this majority is increasing. Similarly, using the same impact measure, we show that the probability of a typical article view being damaged is small but increasing, and we present empirically grounded classes of damage. Finally, we make policy recommendations for Wikipedia and other wikis in light of these findings.nd other wikis in light of these findings.
Added by wikilit teamYes +
Collected data time dimensionLongitudinal +
Comments"We show that 1/10th of 1% of editors cont"We show that 1/10th of 1% of editors contributed nearly half of the value, measured by words read... [and] 42% of damage is repaired almost

immediately, i.e., before it can confuse, offend, or mislead anyone." "[B]ecause a very small proportion of Wikipedia editors account for most of its value, it is important to keep them happy, for example by ensuring that they gain appropriate visibility and status. However, turnover is inevitable in any online community. Wikipedia should also develop policies, tools, and user interfaces to bring in newcomers, teach them community norms, and help them become effective editors." p.267-268 Wikipedia pages; request logs." p.267-268

Wikipedia pages; request logs
ConclusionOur view-based metrics let us both sharpenOur view-based metrics let us both sharpen previous results

and go beyond them. Others have shown that 1% of Wikipedia editors contributed about half of edits [6]. We show that 1/10th of 1% of editors contributed nearly half of the value, measured by words read. Others have shown that one type of damage was repaired quickly [20]. We show this for all types of damage. We also show what this result means for readers: 42% of damage is repaired almost immediately, i.e., before it can confuse, offend, or mislead anyone. Nonetheless, there are still hundreds of millions of damaged views. We categorize the types of damage that occured, show how often they occured, describe their potential impact on readers, and discuss how hard (or easy) they are to detect automatically. We give examples of especially impactful damage to illustrate these points. Finally, we show that the probability of encountering damage increased exponentially from January 2003 to June 2006. What are the implications of our results? First, because a very small proportion of Wikipedia editors account for most of its value, it is important to keep them happy, for example by ensuring that they gain appropriate visibility and status. However, turnover is inevitable in any online community. Wikipedia should also develop policies, tools, and user interfaces to bring in newcomers, teach them community norms, and help them become effective editors. Second, we speculate that the exponential increase in the probability of encountering damage was stopped by the widespread use of anti-vandalism bots. It is likely that vandals will continue working to defeat the bots, leading to an arms race. Thus, continued work on automatic detection of damage is important. Our results suggest types of damage to focus on; the good news is that the results show little subtlety among most vandals. We also generally believe in augmentation, not automation. That is, we prefer intelligent task routing [7] approaches, where automation directs humans to potential damage incidents, but humans make the final decision.dents, but humans make

the final decision.
Conference locationSanibel Island, FL, United states +
Data sourceComputer usage logs + and Wikipedia pages +
Dates4-7 +
Doi10.1145/1316624.1316663 +
Google scholar urlhttp://scholar.google.com/scholar?ie=UTF-8&q=%22Creating%2C%2Bdestroying%2C%2Band%2Brestoring%2Bvalue%2Bin%2BWikipedia%22 +
Has authorReid Priedhorsky +, Jilin Chen +, Shyong Tony K. Lam +, Katherine Panciera +, Loren Terveen + and John Riedl +
Has domainInformation systems +
Has topicOther content topics +, Vandalism +, Participation trends + and Collaboration software +
MonthNovember +
Pages259-268 +
Peer reviewedYes +
Publication typeConference paper +
Published inGROUP '07 Proceedings of the 2007 international ACM conference on Supporting group work +
PublisherAssociation for Computing Machinery +
Research designStatistical analysis +
Research questionsWe pose three specific research questions:We pose three specific research questions:

1. Creating value: Who contributes Wikipedia’s value? Is it the handful of people who edit thousands of times, or is it the thousands of people who edit a handful of times? 2. Impact of damage: What is the impact of damage such as nonsensical, offensive, or false content? How quickly is it repaired, and how much of it persists long enough to confuse, offend, or mislead readers? 3. Types of damage: What types of damage occur, and how often?What types of damage occur, and

how often?
Revid10,717 +
TheoriesUndetermined
Theory typeAnalysis +
TitleCreating, destroying, and restoring value in Wikipedia
Unit of analysisArticle view + and Edit +
Urlhttp://dl.acm.org/citation.cfm?id=1316663 +
Wikipedia coverageMain topic +
Wikipedia data extractionDump +
Wikipedia languageEnglish +
Wikipedia page typeArticle +
Year2007 +