Redigerer Wikipedia (avsnitt)

=== Bruk i forskning ===

Wikipedias [[tekstkorpus]] har blitt bredt brukt for forskning i databasert [[lingvistikk]] (engelsk: Computational linguistics), [[informasjonsgjenfinning]] og prosessering av naturlige språk (engelsk: Natural language processing). Wikipedias tekst blir særlig brukt som en kunnskapsbase for problemet med kobling av navn (engelsk: entity linking) - at Paris betyr byen i Frankrike, og ikke kvinnen [[Paris Hilton]], eller noe annet - som da kalles «wikifisering»,<ref name="wikify">{{cite conference |url=https://www.cse.unt.edu/~tarau/teaching/NLP/papers/Mihalcea-2007-Wikify-Linking_Documents_to_Encyclopedic.pdf |title=Wikify!: linking documents to encyclopedic knowledge |first1=Mihalcea |last1=Rada |first2=Andras |last2=Csomai |author1-link=Rada Mihalcea |date=november 2007 |conference=ACM Conference on Information and Knowledge Management |book-title=CIKM '07: Proceedings of the sixteenth ACM conference on Conference on information and knowledge management |publisher=[[Association for Computing Machinery]] |archive-url=https://web.archive.org/web/20160218062051/https://www.cse.unt.edu/~tarau/teaching/NLP/papers/Mihalcea-2007-Wikify-Linking_Documents_to_Encyclopedic.pdf |archive-date=18. februar 2016 |location=Lisbon; New York City, NY |pages=233–242 |isbn=978-1-59593-803-9 |doi=10.1145/1321440.1321475 |url-status=live}}</ref> og til det relaterte problemet med Word-sense disambiguation.<ref name="milne witten WP usage 1">{{cite conference |url=https://www.cms.waikato.ac.nz/~ihw/papers/08-DNM-IHW-LearningToLinkWithWikipedia.pdf |title=Learning to Link with Wikipedia |first1=David |last1=Milne |first2=Ian H. |last2=Witten |author2-link=Ian H. Witten |date=oktober 2008 |conference=ACM Conference on Information and Knowledge Management |book-title=CIKM '08: Proceedings of the seventeenth ACM conference on Conference on information and knowledge management |publisher=Association for Computing Machinery |location=Napa Valley, CA; New York City, NY |pages=509–518 |isbn=978-1-59593-991-3 |doi=10.1145/1458082.1458150}}</ref> Metoder som tilsvarer wikifisering kan igjen bli brukt for å finne «manglende» lenker i Wikipedia.<ref name="discovering missing WP links 1">{{cite conference |url=https://staff.science.uva.nl/~mdr/Publications/Files/linkkdd2005.pdf |title=Discovering missing links in Wikipedia |author1=Sisay Fissaha Adafre |author2=Maarten de Rijke |author2-link=Maarten de Rijke |date=august 2005 |conference=ACM LinkKDD |book-title=LinkKDD '05: Proceedings of the 3rd international workshop on Link discovery |publisher=Association for Computing Machinery |archive-url=https://web.archive.org/web/20120717054413/https://staff.science.uva.nl/~mdr/Publications/Files/linkkdd2005.pdf |archive-date=17, juli 2012 |location=Chicago, IL; New York City, NY |pages=90–97 |isbn=978-1-59593-135-1 |doi=10.1145/1134271.1134284 |url-status=live}}</ref>

I 2015 publiserte de franske forskerne, José Lages ved Université de Franche-Comté i [[Besançon]], og Dima Shepelyansky ved [[Université Toulouse-III-Paul-Sabatier]] i [[Toulouse]], en global rangering av universiteter, basert på faglige sitater i Wikipedia.<ref name=mitmining>{{cite news |title = Wikipedia-Mining Algorithm Reveals World's Most Influential Universities: An algorithm's list of the most influential universities contains some surprising entries |url = https://www.technologyreview.com/view/544266/wikipedia-mining-algorithm-reveals-worlds-most-influential-universities/ |access-date = 27. desember 2015 |work = MIT Technology Review |date = 7. desember 2015 |accessdate = 2021-10-29 |archivedate = 2016-02-01 |archiveurl = https://web.archive.org/web/20160201174817/https://www.technologyreview.com/view/544266/wikipedia-mining-algorithm-reveals-worlds-most-influential-universities/ |url-status = yes }}</ref><ref name=harvardisonlymarmow>{{cite news |last1 = Marmow Shaw |first1 = Jessica |title = Harvard is only the 3rd most influential university in the world, according to this list |url = https://www.marketwatch.com/story/two-universities-beat-harvard-in-this-surprising-school-ranking-2015-12-09 |access-date = 27. desember 2015 |work = MarketWatch |date = 10. desember 2015}}</ref><ref name=wikipediarankingtimesworldunifranche>{{cite news |last1 = Bothwell |first1 = Ellie |title = Wikipedia Ranking of World Universities: the top 100. List ranks institutions by search engine results and Wikipedia appearances |url = https://www.timeshighereducation.com/features/wikipedia-ranking-world-universities-top-100 |access-date = 27. desember 2015 |work = [[Times Higher Education]] |date = 15. desember 2015}}</ref> De benyttet [[PageRank]], CheiRank og tilsvarende [[algoritme]]r «fulgt av antallet forekomster i de 24 forskjellige språkversjonene av Wikipedia (synkende rekkefølge) og århundret når de ble funnet (stigende rekkefølge).
<ref group="note">"followed by the number of appearances in the 24 different language editions of Wikipedia (descending order) and the century in which they were founded (ascending order)".</ref><ref name=wikipediarankingtimesworldunifranche /><ref>{{cite journal | title = Wikipedia ranking of world universities | author1 = Lages, J. | author2 = Patt, A. | author3 = Shepelyansky, D. | journal = Eur. Phys. J. B | volume = 89 | number = 69 | year = 2016 | page = 69 | doi = 10.1140/epjb/e2016-60922-0| arxiv = 1511.09021 | bibcode = 2016EPJB...89...69L | s2cid = 1965378 }}</ref> Undersøkelsen ble oppdatert i 2019.<ref>{{cite journal | author1 = Coquidé, C. | author2 = Lages, J. | author3 = Shepelyansky, D.L. | title = World influence and interactions of universities from Wikipedia networks. | journal = Eur. Phys. J. B | volume = 92 | number = 3 | year = 2019 | page = 3 | doi = 10.1140/epjb/e2018-90532-7| arxiv = 1809.00332 | bibcode = 2019EPJB...92....3C | s2cid = 52154548 }}</ref>

En studie fra det amerikanske universitetet [[Massachusetts Institute of Technology]] i 2017 antyder at ord brukt i artikler på Wikipedia, ender opp i vitenskapelige publikasjoner.<ref>{{cite news |last1=Brookshire |first1=Bethany |title=Wikipedia has become a science reference source even though scientists don't cite it |url=https://www.sciencenews.org/blog/scicurious/wikipedia-science-reference-citations |work=ScienceNews |date=5. februar 2018 |department=SciCurious |url-access=subscription |url-status=live |archive-date=10. februar 2018 |archive-url=https://web.archive.org/web/20180210120955/https://www.sciencenews.org/blog/scicurious/wikipedia-science-reference-citations}}</ref><ref>{{cite news| first1=Neil C. |last1=Thompson |first2=Douglas |last2=Hanley |title=Science Is Shaped by Wikipedia: Evidence From a Randomized Control Trial |date=2017}}</ref>

Undersøkelser knyttet til Wikipedia har brukt [[maskinlæring]] og [[kunstig intelligens]] for å støtte ulike operasjoner. En av de viktigste områdene er automatisk oppdagelse av vandalisme,<ref>{{cite conference |last1=Sarabadani |first1=Amir |last2=Halfaker |first2=Aaron |last3=Taraborelli |first3=Dario |author2-link=Aaron Halfaker |title=Building automated vandalism detection tools for Wikidata |date=april 2017 |conference= International Conference on World Wide Web Companion |book-title=WWW '17 Companion: Proceedings of the 26th International Conference on World Wide Web Companion |publisher=Association for Computing Machinery |location=Perth, Australia; New York City, NY |pages=1647–1654 |isbn=978-1-4503-4914-7 |doi=10.1145/3041021.3053366 |arxiv=1703.03861}}</ref><ref>{{cite conference |last1=Potthast |first1=Martin |last2=Stein |first2=Benno |last3=Gerling |first3=Robert |title=Automatic Vandalism Detection in Wikipedia |book-title=Advances in Information Retrieval |date=2008 |volume=4956 |pages=663–668 |doi=10.1007/978-3-540-78646-7_75 |series=Lecture Notes in Computer Science |isbn=978-3-540-78645-0 |editor1-first=Craig |editor1-last=Macdonald |editor2-first=Iadh |editor2-last=Ounis |editor3-first=Vassilis |editor3-last=Plachouras |editor4-first=Ian |editor4-last=Ruthven |editor5-first=Ryen W. |editor5-last=White |conference=30th ECIR |location=Glasgow, UK |publisher=Springer}}</ref> og vurdering av [[datakvalitet]] i Wikipedia.<ref>{{cite journal |last1=Asthana |first1=Sumit |last2=Halfaker |first2=Aaron |editor1-last=Lampe |editor1-first=Cliff |title=With Few Eyes, All Hoaxes are Deep |journal=Proceedings of the ACM on Human-Computer Interaction |date=november 2018 |volume=2 |issue=CSCW |doi=10.1145/3274290 |at=21 |publisher=Association for Computing Machinery |location=New York City, NY |issn=2573-0142 |doi-access=free}}</ref>