conText provides a fast, flexible and transparent framework to estimate context-specific word and short document embeddings using the ‘a la carte’ embeddings approach developed by Khodak et al. (2018) and evaluate hypotheses about covariate effects on embeddings using the regression framework developed by Rodriguez et al. (2021).
install.packages("conText")
To use conText you will need three objects:
conText includes sample objects for all three but keep in mind these are just meant to illustrate function implementations. In this Dropbox folder we have included the raw versions of these objects including the full Stanford GloVe 300-dimensional embeddings (labeled glove.rds) and its corresponding transformation matrix estimated by Khodak et al. (2018) (labeled khodakA.rds). We provide an equivalent RDS file for the 2024 GloVe embeddings released in July 2025 (labeled _glove_2024.rds).
Check out this Quick
Start Guide to get going with conText
(last updated:
07/28/2025).
As noted in Rodriguez
et al. (2023) (p. 1272), distance measures typically used to compare
representations in high-dimensional space (such as embedding vectors)
exhibit statistical bias. In Green
et al. (2025), we explore the severity of this problem for
text-as-data applications and provide and validate a bias correction for
the squared Euclidean distance. We implement this estimator and other
recommendations from the paper in the latest update to the
conText()
function. Please refer to the Bias
in Distance Measures vignette for additional information and the Quick
Start Guide for examples of how to use the new version of the
function and a description of changes in the output.
For those working in languages other than English, we have a set of data and code resources here