I made a R wrapper for the Gallicagram API that provides direct access to a set of French corpora and enables to easily run NLP analyses on these corpora.
rallicagram
is a R wrapper for the Gallicagram API. More info can be found on the rallicagram
website and in particular in its main vignette.
In a nutshell, it enables to easily and quickly run some basic Natural Language Processing analyzes on a set of French corpora.
For instance, I use rallicagram
to analyze discourses around the environment in France. It enables to recover the newspaper coverage of a specific term in a few lines of code. Here, “climatique” (a rough proxy for “climate change”), in the Le Monde newspaper, between 1945 and today:
rallicagram::gallicagram(keyword = "climatique")
Here is the plot of the time series obtained:
This package also allows to run more thorough analyzes. For instance, it enables to study the context in which keywords are used. Here is an example from my project on environmental representations:
More info on Gallicagram and its API can be found on a preprint by Gallicagram developers Benoît de Courson and Benjamin Azoulay, on the “Notice” tab of the Gallicagram website and on the API documentation. I built the wrapper through discussions with Benoît and Benjamin. It enabled us to develop a few additional API routes in order to offer to users the possibility of running more thorough NLP analyzes directly through the API.