MasakhaNER: Named entity recognition for African languages
David Ifeoluwa Adelani
(1, 2)
,
Jade Abbott
(3, 2)
,
Graham Neubig
(4)
,
Daniel d'Souza
(5, 2)
,
Julia Kreutzer
(6, 2)
,
Constantine Lignos
(7, 2)
,
Chester Palen-Michel
(7, 2)
,
Happy Buzaaba
(8, 2)
,
Shruti Rijhwani
(4)
,
Sebastian Ruder
(9)
,
Stephen Mayhew
(10)
,
Israel Abebe Azime
(11, 2)
,
Shamsuddeen H Muhammad
(12, 13, 2)
,
Chris Chinenye Emezue
(14, 2)
,
Joyce Nakatumba-Nabende
(15, 2)
,
Perez Ogayo
(16, 2)
,
Anuoluwapo Aremu
(17, 2)
,
Catherine Gitau
(2)
,
Derguene Mbaye
(2)
,
Jesujoba Alabi
(18, 2)
,
Seid Muhie Yimam
(19)
,
Tajuddeen Rabiu Gwadabe
(20, 2)
,
Ignatius Ezeani
(21, 2)
,
Rubungo Andre Niyongabo
(22, 2)
,
Jonathan Mukiibi
(15)
,
Verrah Otiende
(23, 2)
,
Iroro Orife
(24)
,
Davis David
(2)
,
Samba Ngom
(2)
,
Tosin Adewumi
(25, 2)
,
Paul Rayson
(21)
,
Mofetoluwa Adeyemi
(2)
,
Gerald Muriuki
(15)
,
Emmanuel Anebi
(2)
,
Chiamaka Chukwuneke
(21)
,
Nkiruka Odu
(26)
,
Eric Peter Wairagala
(15)
,
Samuel Oyerinde
(2)
,
Clemencia Siro
(2)
,
Tobius Saul Bateesa
(15)
,
Temilola Oloyede
(2)
,
Yvonne Wambui
(2)
,
Victor Akinode
(2)
,
Deborah Nabagereka
(15)
,
Maurice Katusiime
(15)
,
Ayodele Awokoya
(27, 2)
,
Mouhamadane Mboup
(2)
,
Dibora Gebreyohannes
(2)
,
Henok Tilaye
(2)
,
Kelechi Nwaike
(2)
,
Degaga Wolde
(2)
,
Abdoulaye Faye
(2)
,
Blessing Sibanda
(28, 2)
,
Orevaoghene Ahia
(29, 2)
,
Bonaventure F P Dossou
(30, 2)
,
Kelechi Ogueji
(31, 2)
,
Ibrahima Thierno
(2)
,
Abdoulaye Diallo
(2)
,
Adewale Akinfaderin
(2)
,
Tendai Marengereke
(2)
,
Salomey Osei
(11, 2)
1
Universität des Saarlandes [Saarbrücken] = Saarland University [Saarbrücken]
2 Masakhane NLP
3 Retro Rabbit
4 CMU - Carnegie Mellon University [Pittsburgh]
5 ProQuest
6 Google Research
7 Brandeis University
8 Université de Tsukuba = University of Tsukuba
9 DeepMind [London]
10 Duolingo
11 AIMS - African Institute for Mathematical Sciences
12 Universidade do Porto = University of Porto
13 BUK - Bayero University Kano
14 TUM - Technische Universität Munchen - Technical University Munich - Université Technique de Munich
15 MAK - Makerere University [Kampala, Ouganda]
16 African Leadership University
17 University of Lagos
18 Max Planck Institute for Informatics [Saarbrücken]
19 UHH - Universität Hamburg
20 UCAS - University of Chinese Academy of Sciences [Beijing]
21 Lancaster University
22 UESTC - University of Electronic Science and Technology of China [Chengdu]
23 United States International University - Africa
24 Niger-Volta Language Technologies Institute
25 LUT - Luleå University of Technology = Luleå Tekniska Universitet
26 AUST - African University of Science and Technology
27 University of Ibadan
28 NUST - Namibia University of Science and Technology
29 InstaDeep
30 Jacobs University = Constructor University [Bremen]
31 University of Waterloo [Waterloo]
2 Masakhane NLP
3 Retro Rabbit
4 CMU - Carnegie Mellon University [Pittsburgh]
5 ProQuest
6 Google Research
7 Brandeis University
8 Université de Tsukuba = University of Tsukuba
9 DeepMind [London]
10 Duolingo
11 AIMS - African Institute for Mathematical Sciences
12 Universidade do Porto = University of Porto
13 BUK - Bayero University Kano
14 TUM - Technische Universität Munchen - Technical University Munich - Université Technique de Munich
15 MAK - Makerere University [Kampala, Ouganda]
16 African Leadership University
17 University of Lagos
18 Max Planck Institute for Informatics [Saarbrücken]
19 UHH - Universität Hamburg
20 UCAS - University of Chinese Academy of Sciences [Beijing]
21 Lancaster University
22 UESTC - University of Electronic Science and Technology of China [Chengdu]
23 United States International University - Africa
24 Niger-Volta Language Technologies Institute
25 LUT - Luleå University of Technology = Luleå Tekniska Universitet
26 AUST - African University of Science and Technology
27 University of Ibadan
28 NUST - Namibia University of Science and Technology
29 InstaDeep
30 Jacobs University = Constructor University [Bremen]
31 University of Waterloo [Waterloo]
Résumé
We take a step towards addressing the underrepresentation of the African continent in NLP research by bringing together different stakeholders to create the first large, publicly available, high-quality dataset for named entity recognition (NER) in ten African languages. We detail the characteristics of these languages to help researchers and practitioners better understand the challenges they pose for NER tasks. We analyze our datasets and conduct an extensive empirical evaluation of stateof-the-art methods across both supervised and transfer learning settings. Finally, we release the data, code, and models to inspire future research on African NLP. 1
Domaines
Informatique et langage [cs.CL]Origine | Fichiers produits par l'(les) auteur(s) |
---|