Datasets:

Modalities:
Text
Formats:
parquet
ArXiv:
Libraries:
Datasets
Dask
License:
cointegrated commited on
Commit
cd72983
·
1 Parent(s): e821605

update changelog, bibliography and readme with Kyrgyz references

Browse files
Files changed (3) hide show
  1. CHANGELOG.md +3 -0
  2. README.md +1 -0
  3. bibliography.bib +26 -0
CHANGELOG.md CHANGED
@@ -8,6 +8,9 @@ In particular:
8
  1. Any updates to pre-existing files shall entail a major version bump.
9
  2. The inclusion of a new language, without changes to other pre-existing data, shall only entail a minor version bump.
10
 
 
 
 
11
  ## [2.0] - 2024-11-08
12
 
13
  ### Added
 
8
  1. Any updates to pre-existing files shall entail a major version bump.
9
  2. The inclusion of a new language, without changes to other pre-existing data, shall only entail a minor version bump.
10
 
11
+ ## [3.0] - 2025-11-03
12
+ - Added `kyr_Cyrl`
13
+
14
  ## [2.0] - 2024-11-08
15
 
16
  ### Added
README.md CHANGED
@@ -136,6 +136,7 @@ See [CHANGELOG.md](CHANGELOG.md) for information about the latest changes.
136
  | `ita` | `Latn` | `ital1282` | Italian | |
137
  | `kas` | `Arab` | `kash1277` | Kashmiri (Arabic script) | |
138
  | `kas` | `Deva` | `kash1277` | Kashmiri (Devanagari script) | |
 
139
  | `knc` | `Arab` | `cent2050` | Central Kanuri (Arabic script) | |
140
  | `knc` | `Latn` | `cent2050` | Central Kanuri (Latin script) | |
141
  | `lij` | `Latn` | `geno1240` | Ligurian (Genoese) | |
 
136
  | `ita` | `Latn` | `ital1282` | Italian | |
137
  | `kas` | `Arab` | `kash1277` | Kashmiri (Arabic script) | |
138
  | `kas` | `Deva` | `kash1277` | Kashmiri (Devanagari script) | |
139
+ | `kir` | `Cyrl` | `kirg1245` | Kyrgyz | |
140
  | `knc` | `Arab` | `cent2050` | Central Kanuri (Arabic script) | |
141
  | `knc` | `Latn` | `cent2050` | Central Kanuri (Latin script) | |
142
  | `lij` | `Latn` | `geno1240` | Ligurian (Genoese) | |
bibliography.bib CHANGED
@@ -81,3 +81,29 @@
81
  address = "Miami, USA",
82
  publisher = "Association for Computational Linguistics"
83
  }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
81
  address = "Miami, USA",
82
  publisher = "Association for Computational Linguistics"
83
  }
84
+
85
+ % for Kyrgyz
86
+ @inproceedings{jumashev-etal-2025-kyrgyz,
87
+ title = "The {K}yrgyz Seed Dataset Submission to the {WMT}25 Open Language Data Initiative Shared Task",
88
+ author = "Jumashev, Murat and
89
+ Tillabaeva, Alina and
90
+ Kasieva, Aida and
91
+ Omurkanov, Turgunbek and
92
+ Musaeva, Akylai and
93
+ Emil Kyzy, Meerim and
94
+ Chagataeva, Gulaiym and
95
+ Washington, Jonathan",
96
+ editor = "Haddow, Barry and
97
+ Kocmi, Tom and
98
+ Koehn, Philipp and
99
+ Monz, Christof",
100
+ booktitle = "Proceedings of the Tenth Conference on Machine Translation",
101
+ month = nov,
102
+ year = "2025",
103
+ address = "Suzhou, China",
104
+ publisher = "Association for Computational Linguistics",
105
+ url = "https://aclanthology.org/2025.wmt-1.84/",
106
+ pages = "1088--1102",
107
+ ISBN = "979-8-89176-341-8",
108
+ abstract = "We present a Kyrgyz language seed dataset as part of our contribution to the WMT25 Open Language Data Initiative (OLDI) shared task. This paper details the process of collecting and curating English{--}Kyrgyz translations, highlighting the main challenges encountered in translating into a morphologically rich, low-resource language. We demonstrate the quality of the dataset through fine-tuning experiments, showing consistent improvements in machine translation performance across multiple models. Comparisons with bilingual and MNMT Kyrgyz-English baselines reveal that, for some models, our dataset enables performance surpassing pretrained baselines in both English{--}Kyrgyz and Kyrgyz{--}English translation directions. These results validate the dataset{'}s utility and suggest that it can serve as a valuable resource for the Kyrgyz MT community and other related low-resource languages."
109
+ }