Jump to content

Visual Analytics for Sustainability and Climate Change/Dariusz Jemielniak

Add links
From Meta, a Wikimedia project coordination wiki
VIZWP Project Activities Community Articles Requirements Tool DMP Report Credits

Interview with Dariusz Jemielniak by Iolanda Pensa and Chiara Somajni, 3 September 2025, CC BY 4.0.

Dariusz Jemielniak is a wikipedian, a full professor of Management at Kozminski University in Poland, where he heads the MINDS (Management in Networked and Digital Societies) department, and an Associate Faculty at the Berkman-Klein Center for Internet and Society at Harvard University. He has served on the Wikimedia Foundation Board of Trustees for a decade (2015-2024), and has been appointed to join the governing board of EIT (European Institute of Innovation and Technology) by the European Council. He also serves as an elected vice-president of the Polish Academy of Sciences. He is the author of Common Knowledge?: An Ethnography of Wikipedia (2014, Stanford University Press) and Thick Big Data. Doing Digital Social Sciences (2020, Oxford University Press), winner of the Dorothy Lee Award for Outstanding Scholarship in the Ecology of Culture in 2015. He has established the largest online dictionary in Poland, ling.pl. His research focuses on open collaboration, peer production, sharing economy and the propagation of disinformation.


Notes from the interview

Profile and engagement with Wikimedia projects

  • Professor of management at Kozminski University in Warsaw, Poland
  • Faculty associate at Berkman Klein Center for Internet Society at Harvard
  • Vice-president of the Polish Academy of Sciences
  • Involved with Wikipedia and Wikimedia projects for about 20 years
  • Editing Wikipedia since 2006, recently studying it
  • Author of the book Common Knowledge? An Ethnography of Wikipedia, Stanford University Press, Stanford 2014, ISBN 978-0-8047-8944-8
  • Member of the Wikimedia Foundation Board of Trustees from 2015 to 2025
  • Researches how people create knowledge, cooperate and spread information; how people collaborate on article editing
  • Research foci: climate change denialism online (mainly on Reddit, X, YouTube and Telegram) and, overall, propagation of anti-scientific views – including how they affect Wikipedia.


Languages

  • Works on Wikipedia by conducting qualitative analysis in English and Polish and quantitative analysis also in German, Italian, French, Russian, Swedish, Dutch and a few more languages.


Quality assessment of Wikipedia articles

  • Conducted research comparing different Wikipedias: quality standards appear to differ significantly amongst Wikipedias, as comparison between featured articles shows: e.g. they differ in construction, proportion of images, references.
  • Presentation by Jemielniak and Włodzimierz Lewoniewski at Wikimania 2025: Analysis of Wikipedia articles on climate change. Lewoniewski developed an algorithm to study complex markers for quality.
  • Best indicator of quality is persistence of text strings: if a text is added to Wikipedia and stays there for a long time, it's usually a marker of high-quality input. Revert wars, blocking vandalism, and administrative actions are indicative that something wrong is happening with the page.
  • People are getting smarter, including climate denialists: they plant links to legitimate studies, which do not necessarily say what the text they are linked to says. So it's becoming increasingly difficult to detect vandalism. Notable examples: Scientology (although the Scientology Church is currently banned from editing Wikipedia), abortion, climate and homoeopathy.


Climate change

  • Jemielniak's team analysed climate change on social networks (Reddit), and sentiment associated to it. By studying several billion discussions the team discovered that people stopped talking about global warming and started talking about climate change about a year after the Republican campaign to change the wording started. Another research topic: how people refer to packaging, why people seem to think that glass packages are friendly for the ecology (they are not, just healthier).
  • Jemielniak's team also studied the Wiki Project Cyclones, which according to their measures, is the most successful collaborative project on Wiki. Overall, people are quick to report on disasters and thoroughly link to quality sources. The team's hypothesis is because it is an immediate phenomenon, people can easily relate to it, and there is no scientific knowledge challenge that prevents reporting it.
  • A research focus is collaboration and coordination, in particular in the construction of hobbiests' wiki projects like Simpsons (https://simpsonswiki.com/wiki/Main_Page) or the Simpsons Star Trek (https://simpsonswiki.com/wiki/Star_Trek).
  • Jemielniak would like to explore likely coordination of climate change denialism (where nobody will start a wiki project!), starting with accounts that are suspicious or that are known to be borderline anti-scientific, and checking to what extent people are exchanging favors and barnstars, and to what extent people are working together on multiple articles. Through social network analysis, people could be clustered. This could be useful for potential vandalism prevention.
  • Jemielnak is going to analyse the quality of the articles on climate change on Wikipedia using the tool his co-author developed
  • Identifying articles related to climate change is a challenge. The Wiki Project classification is also not robust, and there are variations across languages. 10 languages were analysed, showing a lack of overlap. Articles like Donald Trump are classified as belonging to climate change despite not being climate change-dominant. Need to categorise articles in a way that clarifies whether climate change is the main topic of the article or not. This requires linguistic analysis, like measuring occurrences of words in the article. For Donald Trump, climate change will be mentioned, but it will not be dominant.
  • If you focus on articles strongly related to climate change, you may underestimate the importance of those which are not, but which are probably the most important in terms of views and impact, e.g. Donald Trump or Greta Thunberg. This can be addressed through clustering.


Gaps

  • Using external sources to identify gaps
  • Interwiki can be used to identify articles present on larger projects but absent on smaller ones. Wiki project climate change identifies several thousand articles in English, but the number is much smaller in other Wikipedias.
  • Comparing linguistic editions helps identify gaps (article present or absent, articles cover different aspects, length of the article, comparative analysis of references). Note that at times in some Wikipedias articles are merged: e.g., on many Wikipedias, "global warming" and "climate change" are one article, not on the English Wikipedia. This could be highlighted as a missing article, but it is not.
  • To editors, it would be useful to compare the lengths of articles, count references and draw references from across projects.


Feedback on prototype

  • Very impressive work. It can be extremely valuable.
  • It would be useful to be able to analyse interest over time of articles in a specific timeframe and compare them; articles could be selected based on shared category, range or by other criteria. E.g.: Ciechanowski K, Banasik-Jemielniak N, Jemielniak D. What's hot and what's not in lay psychology: Wikipedia's most-viewed articles. Curr Psychol. 2022 Oct 12:1-13. doi: 10.1007/s12144-022-03826-0. Epub ahead of print. PMID: 36248218; PMCID: PMC9553632
  • Categories are a lame means to create a subset: consider the possibility of attributing names to articles to create a subset of interest that one can autonomously analyse
  • Comparing across four languages is wonderful and very useful


Potential collaborations

  • Wikirank's (wikirank dot net) quality assessment methodology could be integrated into the tool, Jemielniak can establish a connection with the author
  • Tools to be considered for integration and coordination: PagePile and PetScan
  • A set of articles made available by Jemielniak will be used to test the tool.

Requirements suggested

[edit]
  • Persistence of information: stable text strings are indicators of articles' quality
  • Dynamics of collaborations (barnstars, collaboration on multiple articles, clusters of people)
  • Analysis of differences between Wikipedia editions, including references
  • Interest over time
  • Define a subset of articles and analyse it
Retrieved from "https://meta.wikimedia.org/w/index.php?title=Visual_Analytics_for_Sustainability_and_Climate_Change/Dariusz_Jemielniak&oldid=29448110"