Item

BRIGHTER: BRIdging the Gap in Human-Annotated Textual Emotion Recognition Datasets for 28 Languages

Muhammad, Shamsuddeen Hassan
; 
Ousidhoum, Nedjma
; 
Abdulmumin, Idris
; 
Wahle, Jan Philip
; 
Ruas, Terry Lima
; 
Beloucif, Meriem
; 
de Kock, Christine
; 
Surange, Nirmal
; 
Teodorescu, Daniela
; 
Ahmad, Ibrahim Said
... show 10 more
Author
Muhammad, Shamsuddeen Hassan
Ousidhoum, Nedjma
Abdulmumin, Idris
Wahle, Jan Philip
Ruas, Terry Lima
Beloucif, Meriem
de Kock, Christine
Surange, Nirmal
Teodorescu, Daniela
Ahmad, Ibrahim Said
Adelani, David Ifeoluwa
Aji, Alham F.
Ali, Felermino Dário Mário António
Alimova, Ilseyar S.
Araujo, Vladimir
Babakov, Nikolay
Baes, Naomi
Bucur, Ana Maria
Bukula, Andiswa
Cao, Guanqun
Tufiño, Rodrigo
Chevi, Rendi
Chukwuneke, Chiamaka
Ciobotaru, Alexandra
Dementieva, Daryna
Gadanya, Murja Sani
Geislinger, Robert
Gipp, Béla
Hourrane, Oumaima
Ignat, Oana
Lawan, Falalu Ibrahim
Mabuya, Rooweither
Mahendra, Rahmad
Marivate, Vukosi N.
Panchenko, Alexander I.
Piper, Andrew
Ferreira, Charles Henrique Porto
Protasov, Vitaly
Rutunda, Samuel
Shrivastava, Manish
Udrea, Aura Cristina
Wanzare, Lilian D.A.
Wu, Sophie
Wunderlich, Florian Valentin
Zhafran, Hanif Muhammad
Zhang, Tianhui
Zhou, Yi
Mohammad, Saif M.
Supervisor
Department
Natural Language Processing
Embargo End Date
Type
Conference proceeding
Date
2025
License
Language
English
Collections
Publications
Show all metadata
Research Projects
Organizational Units
Journal Issue
Abstract
People worldwide use language in subtle and complex ways to express emotions. Although emotion recognition-an umbrella term for several NLP tasks-impacts various applications within NLP and beyond, most work in this area has focused on high-resource languages. This has led to significant disparities in research efforts and proposed solutions, particularly for under-resourced languages, which often lack high-quality annotated datasets. In this paper, we present BRIGHTER-a collection of multi-labeled, emotion-annotated datasets in 28 different languages and across several domains. BRIGHTER primarily covers low-resource languages from Africa, Asia, Eastern Europe, and Latin America, with instances labeled by fluent speakers. We highlight the challenges related to the data collection and annotation processes, and then report experimental results for monolingual and crosslingual multi-label emotion identification, as well as emotion intensity recognition. We analyse the variability in performance across languages and text domains, both with and without the use of LLMs, and show that the BRIGHTER datasets represent a meaningful step towards addressing the gap in text-based emotion recognition.
Citation
S. H. Muhammad et al., “BRIGHTER: BRIdging the Gap in Human-Annotated Textual Emotion Recognition Datasets for 28 Languages,” vol. 1, pp. 8895–8916, Aug. 2025, doi: 10.18653/V1/2025.ACL-LONG.436
Source
Proceedings of the Annual Meeting of the Association for Computational Linguistics
Conference
63rd Annual Meeting of the Association for Computational Linguistics, ACL 2025
Keywords
Emotion Recognition Datasets, Low-Resource Languages, Multilingual NLP, Multi-label Emotion Identification, Emotion Intensity Recognition, Cross-lingual Generalisation, Human-Annotated Textual Data, Domain and Language Diversity
Subjects
Source
63rd Annual Meeting of the Association for Computational Linguistics, ACL 2025
Publisher
Association for Computational Linguistics
DOI
10.18653/v1/2025.acl-long.436
Additional links
https://aclanthology.org/2025.acl-long.436/
Full-text link