On Classification with Large Language Models in Cultural Analytics

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Bamman, David, Chang, Kent K., Lucy, Li, Zhou, Naitian
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914974165630976
author Bamman, David
Chang, Kent K.
Lucy, Li
Zhou, Naitian
author_facet Bamman, David
Chang, Kent K.
Lucy, Li
Zhou, Naitian
contents In this work, we survey the way in which classification is used as a sensemaking practice in cultural analytics, and assess where large language models can fit into this landscape. We identify ten tasks supported by publicly available datasets on which we empirically assess the performance of LLMs compared to traditional supervised methods, and explore the ways in which LLMs can be employed for sensemaking goals beyond mere accuracy. We find that prompt-based LLMs are competitive with traditional supervised models for established tasks, but perform less well on de novo tasks. In addition, LLMs can assist sensemaking by acting as an intermediary input to formal theory testing.
format Preprint
id arxiv_https___arxiv_org_abs_2410_12029
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle On Classification with Large Language Models in Cultural Analytics
Bamman, David
Chang, Kent K.
Lucy, Li
Zhou, Naitian
Computation and Language
Computers and Society
In this work, we survey the way in which classification is used as a sensemaking practice in cultural analytics, and assess where large language models can fit into this landscape. We identify ten tasks supported by publicly available datasets on which we empirically assess the performance of LLMs compared to traditional supervised methods, and explore the ways in which LLMs can be employed for sensemaking goals beyond mere accuracy. We find that prompt-based LLMs are competitive with traditional supervised models for established tasks, but perform less well on de novo tasks. In addition, LLMs can assist sensemaking by acting as an intermediary input to formal theory testing.
title On Classification with Large Language Models in Cultural Analytics
topic Computation and Language
Computers and Society
url https://arxiv.org/abs/2410.12029