Testing Hypotheses of Covariate Effects on Topics of Discourse
Statistical Analysis and Data Mining: An ASA Data Science Journal · 2026-04 · 19 (2)
DOI: 10.1002/sam.70066 ↗Kurz gefasst
Fachartikel in der Zeitschrift Statistical Analysis and Data Mining: An ASA Data Science Journal, online erschienen am 26.02.2026. Verfasst von Gabriel Phelan und 1 weiteren Autor:innen. Die Arbeit nennt 47 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: Natural Sciences and Engineering Research Council of Canada.
Abstract (Originalsprache)
We introduce an approach to topic modeling with document‐level covariates that remains tractable in the face of large text corpora. This is achieved by de‐emphasizing the role of parameter estimation in an underlying probabilistic model, assuming instead that the data come from a fixed but unknown distribution whose statistical functionals are of interest. We propose combining a convex formulation of non‐negative matrix factorization with standard regression techniques as a fast‐to‐compute and useful estimate of such a functional. Uncertainty quantification can then be achieved by reposing non‐parametric resampling methods on top of this scheme. This is in contrast to popular topic modeling paradigms, which posit a complex and often hard‐to‐fit generative model of the data. We argue that the simple, non‐parametric approach advocated here is faster, more interpretable, and enjoys better inferential justification than said generative models. Finally, our methods are demonstrated with an application analyzing covariate effects on discourse of flavors attributed to Canadian beers.
Förderung
- Natural Sciences and Engineering Research Council of Canada
Zitieren
APA
Phelan, G., & Campbell, D. A. (2026). Testing Hypotheses of Covariate Effects on Topics of Discourse. Statistical Analysis and Data Mining: An ASA Data Science Journal, 19(2). https://doi.org/10.1002/sam.70066
Vancouver
Phelan G, Campbell DA. Testing Hypotheses of Covariate Effects on Topics of Discourse. Statistical Analysis and Data Mining: An ASA Data Science Journal. 2026;19(2). doi:10.1002/sam.70066
BibTeX
@article{Phelan2026,
author = {Phelan, Gabriel and Campbell, David A.},
title = {Testing Hypotheses of Covariate Effects on Topics of Discourse},
journal = {Statistical Analysis and Data Mining: An ASA Data Science Journal},
year = {2026},
volume = {19},
number = {2},
doi = {10.1002/sam.70066}
}
EUVIMED