CaLMQA: Exploring culturally specific long-form question answering across 23 languages

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Arora, Shane, Karpinska, Marzena, Chen, Hung-Ting, Bhattacharjee, Ipsita, Iyyer, Mohit, Choi, Eunsol
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908402468257792
author Arora, Shane
Karpinska, Marzena
Chen, Hung-Ting
Bhattacharjee, Ipsita
Iyyer, Mohit
Choi, Eunsol
author_facet Arora, Shane
Karpinska, Marzena
Chen, Hung-Ting
Bhattacharjee, Ipsita
Iyyer, Mohit
Choi, Eunsol
contents Despite rising global usage of large language models (LLMs), their ability to generate long-form answers to culturally specific questions remains unexplored in many languages. To fill this gap, we perform the first study of textual multilingual long-form QA by creating CaLMQA, a dataset of 51.7K culturally specific questions across 23 different languages. We define culturally specific questions as those that refer to concepts unique to one or a few cultures, or have different answers depending on the cultural or regional context. We obtain these questions by crawling naturally-occurring questions from community web forums in high-resource languages, and by hiring native speakers to write questions in under-resourced, rarely-studied languages such as Fijian and Kirundi. Our data collection methodologies are translation-free, enabling the collection of culturally unique questions like "Kuber iki umwami wa mbere w'uburundi yitwa Ntare?" (Kirundi; English translation: "Why was the first king of Burundi called Ntare (Lion)?"). We evaluate factuality, relevance and surface-level quality of LLM-generated long-form answers, finding that (1) for many languages, even the best models make critical surface-level errors (e.g., answering in the wrong language, repetition), especially for low-resource languages; and (2) answers to culturally specific questions contain more factual errors than answers to culturally agnostic questions -- questions that have consistent meaning and answer across many cultures. We release CaLMQA to facilitate future research in cultural and multilingual long-form QA.
format Preprint
id arxiv_https___arxiv_org_abs_2406_17761
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle CaLMQA: Exploring culturally specific long-form question answering across 23 languages
Arora, Shane
Karpinska, Marzena
Chen, Hung-Ting
Bhattacharjee, Ipsita
Iyyer, Mohit
Choi, Eunsol
Computation and Language
Artificial Intelligence
Machine Learning
Despite rising global usage of large language models (LLMs), their ability to generate long-form answers to culturally specific questions remains unexplored in many languages. To fill this gap, we perform the first study of textual multilingual long-form QA by creating CaLMQA, a dataset of 51.7K culturally specific questions across 23 different languages. We define culturally specific questions as those that refer to concepts unique to one or a few cultures, or have different answers depending on the cultural or regional context. We obtain these questions by crawling naturally-occurring questions from community web forums in high-resource languages, and by hiring native speakers to write questions in under-resourced, rarely-studied languages such as Fijian and Kirundi. Our data collection methodologies are translation-free, enabling the collection of culturally unique questions like "Kuber iki umwami wa mbere w'uburundi yitwa Ntare?" (Kirundi; English translation: "Why was the first king of Burundi called Ntare (Lion)?"). We evaluate factuality, relevance and surface-level quality of LLM-generated long-form answers, finding that (1) for many languages, even the best models make critical surface-level errors (e.g., answering in the wrong language, repetition), especially for low-resource languages; and (2) answers to culturally specific questions contain more factual errors than answers to culturally agnostic questions -- questions that have consistent meaning and answer across many cultures. We release CaLMQA to facilitate future research in cultural and multilingual long-form QA.
title CaLMQA: Exploring culturally specific long-form question answering across 23 languages
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2406.17761