Who's the most significant historical figure?

We work in the fields of data and computer science and do not answer these questions as historians might, through a principled assessment of a person's achievements. Instead, we aggregate millions of opinions. We rank historical figures just as Google ranks web pages, by integrating a diverse set of measurements of reputation into a single consensus value.Significance is related to fame but measures something different. According to our system, forgotten US President Chester A Arthur (who we rank at 499) is more historically significant than pop star Justin Bieber (ranked 8,633), even though Arthur may have a less devoted following and certainly has lower contemporary name recognition. We believe our computational, data-centric analysis provides new ways to understand and interpret the past.Historically significant figures leave statistical evidence of their presence behind, if one knows where to look for it. We use several data sources to fuel our ranking algorithms. Most important is Wikipedia, the web-based, collaborative, multi-lingual encyclopedia. Wikipedia is enormous, featuring well over 3m articles in its English edition alone. But we use it in a manner quite different from the typical reader, by analysing the Wiki pages of more than 800,000 people to measure quantities that should correspond to historical significance. We would expect that more significant people should have longer Wikipedia pages than those less notable because they have greater accomplishments to report. The Wiki pages of people of higher significance should attract greater readership than those of lower significance. The elite should have pages linked to by other highly significant figures, meaning they should have a high PageRank, the measure of importance used by Google to identify important web pages. We combine these other variables into a single number using a statistical method called factor analysis. But we need one final correction: to fairly compare contemporary figures such as Britney Spears against, say, Aristotle, we must adjust for the fact that today's stars will fade from living memory over the next several generations. By analysing traces left in millions of scanned books, we hope to measure just how fast this decay occurs, and correct for it.We have naturally received strong reactions from readers of our book Who's Bigger? complaining about our computational methodology. Certain historians have complained that Wiki cannot be trusted as a source for anything. This is pretty silly. People find Wikipedia articles to be generally accurate and informative, or else they wouldn't read them. Where do you head to read up on a new topic you are interested in? We think it is clear that anyone (or anything, like our algorithms) that has read all of Wikipedia would be in an excellent position to discourse about the most important people in recorded history.More cogent is the complaint that our results are culturally biased because we analyse only the English edition of Wikipedia. How can we fairly assess the significance of Chinese poets against US presidents? We agree that any ranking of historical significance is indeed culturally dependent and so, yes, our rankings have an Anglocentric bias. But the depth of Wikipedia is so great that there are hundreds of articles about Chinese poets in the English edition.Others highlight a few contemporary figures that they deem us to have overrated, such as Britney Spears (689) or Barack Obama (111), and use this anecdotal evidence to sneer. But we also conduct validation procedures, and compare our rankings to public opinion polls, Hall of Fame voting records, sports statistics, and even the prices of paintings and autographs.Anecdotal evidence is not as compelling as it might seem. British readers have complained that our algorithms don't rank British figures high enough just as strongly as Spanish readers think we are unfair to their compatriots. But our book is designed in part to generate debate.(theguardian.com)ANN.Az