dhCenter Scientific Committee Profile
Jean-Philippe Cointet, researcher, Sciences Po médialab
A self-described interdisciplinary “hybrid”, Jean-Philippe Cointet is an engineer and computer scientist by training, with a diverse research background in science and technology studies and sociology. He brings his experience with text corpora as well as interdisciplinary research more globally to the dhCenter Scientific Committee.
Jean-Philippe Cointet’s research interests range from social media analysis to political process mapping. He received his PhD in Complex Systems from the Ecole Polytechnique in Paris in 2009, and defended his habilitation in social sciences at the Ecole Normale Supérieure in 2017. Since 2017, he has been a researcher specializing in text analysis and the sociopolitical dynamics of corpora at the médialab of the Paris Institute of Political Studies (Sciences Po). Specifically, he designs innovative computational sociology methods, including original methods for modeling digital traces to map public space and its dynamics.
Cointet coordinates the project GOPI (La géométrie des problèmes publics), funded by the Agence nationale de recherche, which designs new word-embedding methods for the modeling of public issues. He is also affiliated with Columbia University’s INCITE research center.
From computer science to CorText
After his PhD, Cointet worked at the French social science lab LISIS, where he first became interested in visualizing and modeling historical changes in scientific dynamics. It was in this context that he began using press and social media data to monitor societal debates on issues related to science and technology.
Today, Cointet is one of the primary designers of the digital text analysis platform CorText, which uses artificial intelligence to analyze data corpora from the social sciences.
He says that he hopes to bring his expertise in natural language processing (NLP) techniques, machine learning algorithms, and network analysis to the dhCenter.
“I can use my prior experience to address questions asked by sociologists or political scientists when they are confronted with a text corpus, such as: When and how can one mix both competencies when designing and developing such a tool? How can one recruit new contributors who are both users and co-designers of future capacities?”
Cementing interdisciplinary collaborations
As a member of the dhCenter Scientific Committee, Cointet emphasizes the importance of support for interdisciplinary research, noting that while this is an essential component of the digital humanities, it can also be a source of frustration.
“Even where there is a genuine desire to collaborate, the practices and norms in the social sciences and humanities and in engineering and computer science are so distant that it can take longer to get on the same page and align expectations,” he says.
“My experience is that these kinds of collaborations can take quite a long time to cement into fruitful interactions. From the dhCenter’s perspective, such initiatives should be given support over a significant period of time, so that new teams can build a common understanding.”