Judging by Wikipedia popularity, who are the most important historical figures of all time? According to one algorithm, it’s 18th-century scientist Carl Linnaeus, followed by Jesus. According to a modified version of the algorithm, it’s Hitler, then Michael Jackson.
In fact, the full lists of top influential figures compared among 24 different language editions of Wikipedia differ substantially, according to a recent study. The one thing they had in common? The top figures, averaged across all lists, were largely post-17th-century Western white guys. Women only made up an average 5% to 10% of the top 100 rankings.
The research team, led by Young-Ho Eom of the University of Toulouse, used two popular page-ranking algorithms to determine the most important or famous historical figures on Wikipedia in different languages.
Eom says that the goal of the research, available on the pre-print server arXiv, wasn’t necessarily to highlight winners of the Wikipedia influence wars, but to measure similarities and differences across cultures. Still, he was surprised to find out that the top rankings on the lists, regardless of origin, often looked the same.
“All Wikipedia editions have strong locality,” Eom explains. “In Korean or Chinese Wikipedia, most of [the important historical figures] are actually Korean or Chinese, but the very top of them, they have a very similar structure. Most of them are not from outside of Europe.”
The algorithms themselves also generated significant differences. The first algorithm the researchers used was PageRank, a tool developed by Google to rank search engine results based on the number of links to a web page. The second algorithm they used, called 2DRank, looked at linking as more of a two-way street: It measured outgoing links, or how many external sources the Wikipedia pages cited as well. The different methods were enough to switch second place from Jesus to Michael Jackson.
The quirky results reveal important biases in the algorithms. Eom thinks that Carl Linnaeus may have topped the PageRank list because he invented the scientific taxonomy for classifying plants and animals. Scientific articles that mention these plants and animals often cite him, which is why his name would be highlighted by a system that values referential linking. Musicians and artists, on the other hand, were much more popular on the 2DRank list. The reason, Eom explains, could be that their extensive citations at the bottom of the page, noting important interviews and other coverage, increased their value in an algorithm that prizes outgoing links.
But neither of these fully explain the European, male, post-Enlightenment bias. Part of the reason could be that Wikipedians skew male and Western. Another could be something of an intrinsic “rich get richer” quality to Wikipedia. Network scientists call it “preferential attachment”–the process by which popular pages only become more popular, while lesser known pages fade into obscurity. Eom says that scientific papers are almost certainly impacted by preferential treatment; as more people cite a paper, it’s only more likely to be cited by others. Eom and his colleagues have not yet seen whether Wikipedia maintains a kind of preferential attachment bias today, but the possibility does exist.
To explore more of the rankings, head here.