#topicmodelling — Public Fediverse posts

It is quite fun to occasionally come back to using #slurm to send commands to a super computer node.

Let's see if running a #TopicModelling script with a V100 GPU reduces the running time from 80+ hours to a few minutes, as I expect :)

It is quite fun to occasionally come back to using #slurm to send commands to a super computer node.

Let's see if running a #TopicModelling script with a V100 GPU reduces the running time from 80+ hours to a few minutes, as I expect :)

Published at #IRRJ: "Exploring Embedding Interpretability by Correspondences Between Topic Models and Text Embeddings" by Meng Yuan, Lida Rashidi, and Justin Zobel. #InformationRetrieval, #EmbeddingInterpretability, #Explanability, #TopicModelling

https://doi.org/10.54195/irrj.23703

Bien que #Rstats soit le cousin pauvre de #Python en traitement de texte (#NLP), rien n'empêche de l'utiliser pour la modélisation thématique (#topicmodelling).

Dans cet exemple, je montre aussi comment soumettre le résultat de l'analyse à une #IA textuelle générative basée #LLM.

Corpus de test: littérature française de la seconde moitié du 18e. Quelqu'un m'expliquera peut-être l'étrange mais quantifiable proximité entre Voltaire et le Marquis de Sade...

https://ourednik.info/maps/2024/05/31/topic-modeling-la-modelisation-thematique-avec-r-et-quanteda

Bien que #Rstats soit le cousin pauvre de #Python en traitement de texte (#NLP), rien n'empêche de l'utiliser pour la modélisation thématique (#topicmodelling).

Dans cet exemple, je montre aussi comment soumettre le résultat de l'analyse à une #IA textuelle générative basée #LLM.

Corpus de test: littérature française de la seconde moitié du 18e. Quelqu'un m'expliquera peut-être l'étrange mais quantifiable proximité entre Voltaire et le Marquis de Sade...

https://ourednik.info/maps/2024/05/31/topic-modeling-la-modelisation-thematique-avec-r-et-quanteda

@dajb Though I’m still sore from the overwhelming LLM chatter of the past few months, I did think about them while reading your piece…
Had neat experiences with other AI/ML approaches to classification. (#TopicModelling methods such as #LatentDirichletAllocation.)

For collaborative design, it does sound like nonhuman actors can contribute. Was also thinking about the iterative nature of design approaches. And about fluid categories.
As we say in #LinguisticAnthropology:

> All grammars leak

@dajb Though I’m still sore from the overwhelming LLM chatter of the past few months, I did think about them while reading your piece…
Had neat experiences with other AI/ML approaches to classification. (#TopicModelling methods such as #LatentDirichletAllocation.)

For collaborative design, it does sound like nonhuman actors can contribute. Was also thinking about the iterative nature of design approaches. And about fluid categories.
As we say in #LinguisticAnthropology:

> All grammars leak