Submissions/Hebrew Wikipedia entries in the eyes of experts
|
Presentation Media 30px No slides known
|
- Review no.
156
- Title of the submission
Hebrew Wikipedia entries in the eyes of experts
- Type of submission
presentation
- Author of the submission
Judit Bar-Ilan, Eti Yaari and Shifra Baruchson-Arbib
- E-mail address
barilaj@mail.biu.ac.il
- Country of origin
Israel
- Affiliation, if any (organization, company etc.)
Department of Information Science, Bar-Ilan University
- Personal homepage or blog
http://www.biu.ac.il/faculty/Judit/
- Abstract (please use no less than 300 words to describe your proposal)
We will present the results of an extensive study, where 60 experts (doctoral students and/or researchers with PhD degrees) evaluated four Hebrew Wikipedia articles each in their specific areas of interest. The Wikipedia articles were sent to them as MS-Word files, with the community templates removed. Each expert received one of each of the following types of articles: featured, regular, an article that needs to be rewritten and an article that needs to be extended. They were asked to pick the best and the worst article in their opinion out of the four articles evaluated and to rate each article on several scales: contribution to the potential reader, updatedness, linguistic style, trustworthiness, neutrality, accuracy, coverage and an overall grade for the article. The experts were also asked to express their opinion about each article, while relating to three specific points: factual mistakes, misleading phrasings and topics missing from the article. The textual information was analyzed using the method of content analysis. This resulted in the identification of 26 content categories and 895 classifications. Since we asked the experts to critically evaluate the articles, naturally there were more negative comments (77% of the classifications) than positive ones. Most of the negative comments related to missing topics (for 49% or for 118 out of the 240 articles examined), and topics that existed in the article, but were not extended enough in their opinion (85 articles, 35%). This is not that surprising, since we specifically asked them about missing topics, and an expert usually has more extensive knowledge on the topic than an encyclopedia article is supposed to cover. Of more interest are the major and minor mistakes identified (for 27% and 17% of the examined articles respectively); problematic phrasing (27%); missing links and references (21%) and bias at least in some parts of the article (19%). Only one expert specifically told us that he was going to edit the actual Wikipedia article in order to improve it.
In terms of the overall grade and all the other scores, except for neutrality the group of featured articles scored highest. The average overall grade for the featured articles was 84 out of 100. The best score on neutrality was received by the group of regular articles, but the differences between the groups were negligible. Rather interestingly the group of articles with the community template “extension is needed” scored the second highest marks, followed by the group of regular articles. The one way Anova test showed significant differences between the overall grades of the featured group of articles and the other groups. There were no significant differences between the other three groups. In addition we also tested within subject differences in the overall grade assigned to the articles by running a repeated measures test, here too the differences were significant, because of the higher grades assigned to the featured articles. These findings show, that although the experts were critical about the Wikipedia articles, they agree with the community that the featured articles are the “best” articles in the Hebrew Wikipedia.
- Track
Knowledge and Collaboration
- Will you attend Wikimania if your submission is not accepted?
Most probably
- Slides or further information (optional)
Interested attendees
If you are interested in attending this session, please sign with your username below. This will help reviewers to decide which sessions are of high interest. Sign with four tildes. (~~~~).