Just kidding… Turns out it can barely handle simple transformations. I still have to correct its result and write my own commands for the complex stuff.

@liamsims @ThomasGuillem1 What does the stamp of the Bibliothèque Saint Geneviève look like?

RT @medialab_ScPo: Il y a un 1 an, le projet @DEFACTO_UE interrogeait les français sur le thème #mieuxsinformer. Une nouvelle exploitation…

Q1: Comment s’appelle ce site web ?

RT @Yomguithereal: Ural, the @medialab_ScPo #python library full or url-related heuristics, just hit v1! Now with more deduplication scheme…

Is there reason to prefer something newer, maybe based on transformer models, to do syntactic dependency parsing? Or is Stanza’s older method still one of the top-performing in the field, despite recent developments?

RT @ADHOrg: #CFP for Computational Humanities Research 2023 (see details:

@boogheta Et cette fois, ils n’ont pas communiqué de date non plus. D’ici “a few more days”… C’est comme ça qu’il faut organiser les projets de recherche pluriannuels, jour après jour.

A thoroughly enjoyable first #FOSDEM

@fpietrosanti @medialab_ScPo … and last but not least, Guillaume Plique (presented by @boogheta), “ipysigma: a Jupyter widget for interactive visual network analysis”

@fpietrosanti @medialab_ScPo Armin Pournaki, “Interactive network visualizations as "guided close reading" devices for the social sciences”

@fpietrosanti @medialab_ScPo @BeatriceMazoyer, “PIMMI: Python IMage MIning”

@fpietrosanti @medialab_ScPo Emilien Schultz, “The under-equipped social scientist ?”

Deadline: 15 January (Sunday) Come work with me!

Back in the States! 🇺🇸 And first time at Georgetown 🤩 This little American will be repping the @medialab_ScPo tech team and presenting some of the open-source #DH tools we use for social media research and the study of #misinformation to our US counterparts at @McCourtSchool