Lancaster Summer Schools

For the past several years, I’ve been wanting to attend one of Lancaster University’s summer schools in corpus linguistics. This year, they are opening up several of their sessions for free to anyone who is interested in corpus linguistics. You can be sure that I’ll be attending!

These open session include (list copied and pasted from the Lancaster Summer Schools website):

  • A series of lectures on different topics and applications of corpus linguistics
  • A practical session on using #LancsBox to analyse different sets of language data
  • A symposium on Innovation in corpus linguistics

For anyone else who wants to join me, you can find all the information here: http://wp.lancs.ac.uk/corpussummerschools/open-events/.

I hope to see you there!

R & Python

Adrien Barbaresi présente une façon d’intégrer Python dans R. Voici le début de la publication sur son blogue Bits of Language: corpus linguistics, NLP and text analytics.

Why choose between R and Python?

R is a free software environment for statistical computing and graphics. Together with Python, it is part of the most popular languages among (data) scientists. Although both environments are similar, most people feel they face a choice between the two. The question “R vs Python, What should I learn?” resonates across the Internet. But why choose between them when you can choose both?

The reticulate package provides a comprehensive set of tools for seamless interoperability between Python and R. It basically allows for execution of Python code inside an R session, so that Python packages can be used with minimal adaptations, which is ideal for those who would rather operate from R than having to go back and forth between languages and environments.

Doing Linguistics with a Corpus

Doing Linguistics with a CorpusEgbert, Jesse, Tove Larsson, et Douglas Biber. 2020. Doing Linguistics with a Corpus: Methodological Considerations for the Everyday User (Elements in Corpus Linguistics). Cambridge: Cambridge University Press. doi:10.1017/9781108888790
 
A very well written and useful book. The general idea is that you always should keep in mind your linguistic research question and be aware that results obtained through corpus tools may create layers of abstraction between the data and the researcher. A basic understanding of what the tool does with your data (which changes it applies to them) is required. You should always come back to the original texts after obtaining results created with the help of statistical methods.
 
Some of my favorite quotes:
 
“we should always strive to choose minimally sufficiant statistical methods, meaning that we should choose tests that are no more no less sophisticated than the study design requires. (…) Even simple, seemingly straightforward statistical methods may lead to linguistically questionable conclusions caused by layers of abstraction between the data and the researcher” (p. 40).
 
“Statistical tests cannot replace linguistical analysis; they are, and should remain, tools that should assist the researcher in drawing linguistically valid conclusions” (p. 42).
 
“an ever-widening gap between the everyday user and the language contained in the corpus” (p. 70).
 
“end goal of turning corpus-based data into meaningful linguistic information” (p. 70).
 
[i]nformation is born when data are interpreted” (Stallings, 1989: 2) (p. 71).
 
“The two sides of any corpus-based linguistically meaningful quantitative analysis: the statistical analysis and the qualitative interpretation (…) through consideration of actual texts” (p. 72).

© Photo : evalac

Prise en main de R

Depuis deux séances, les membres du Club Corpus lisent ensemble des chapitres sélectionnés du livre cité ci-dessous pour prendre en main R, selon notre habitude de compagnonnage et entraide compréhensive et amicale. La prochaine séance aura lieu le 16 avril 2021 à 14h, par Zoom. Dans l’idéal, vous aurez lu les chapitres 2 et 3 avant cette date.

Desagulier, Guillaume. 2017. Corpus Linguistics and Statistics with R: Introduction to Quantitative Methods in Linguistics. 1st ed. 2017. Quantitative Methods in the Humanities and Social Sciences. Cham: Springer International Publishing : Imprint: Springer. https://doi.org/10.1007/978-3-319-64572-8.

king – man + woman = queen

…ou comment comprendre ce que c’est qu’un vecteur

Comme déjà précisé dans l’article précédent, les séances actuelles du Club Corpus sont dédiées à la lecture collective de chapitres choisis de Desagulier, Guillaume (2017). Corpus Linguistics and Statistics with R : Introduction to Quantitative Methods in Linguistics (1st ed. 2017). Springer International Publishing : Imprint: Springer. https://doi.org/10.1007/978-3-319-64572-8

Le chapitre 2.9.1 traite des vecteurs. En m’intéressant à ce thème, je suis tombée sur Introduction to Word Vectors, un article de Jayesh Bapu Ahire, chercheur et consultant, publié sur DZone. Intéressant et éclairant ! Cet article m’aide à mieux comprendre certaines fonctionnalités d’outils de corpus comme le thésaurus de Sketch Engine.

Jakubícˇek, M., & Rychlý, P. (2019). A Distributional Multi-word Thesaurus in Sketch Engine. In A. Horák, P. Rychlý, & A. Rambousek (Éds.), Proceedings of Recent Advances in Slavonic Natural Language Processing (p. 143‑147). https://nlp.fi.muni.cz/raslan/raslan19.pdf

 

Lire Pierre Frath

Lors de la séance du Club Corpus du 11 décembre 2020, tenue via Zoom, Pierre Frath nous a fait une présentation sur les apports du corpus à partir de l’exemple de la causalité. Cela a mené à une discussion très intéressante portant, entre autres, sur l’usage de la formule set in.
Le livre de Pierre Frath cité ci-dessous nous permettra de prolonger ces échanges.
 
Frath, Pierre. 2020. Linguistique anthropologique et référentielle. Sapientia Hominis.