Update README.md
Browse files
README.md
CHANGED
|
@@ -9,4 +9,81 @@ tags:
|
|
| 9 |
- yucatec maya
|
| 10 |
- low-resource languages
|
| 11 |
- indigenous languages
|
| 12 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
- yucatec maya
|
| 10 |
- low-resource languages
|
| 11 |
- indigenous languages
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
# maya2vec
|
| 15 |
+
|
| 16 |
+
maya2vec is a model to encode word embeddings in Maya.
|
| 17 |
+
|
| 18 |
+

|
| 19 |
+
|
| 20 |
+
|
| 21 |
+
maya2vec embeddings use 512 dimensions and were trained using the Skip-gram with Negative Sampling algorithm (SGNS) on data from La Jornada Maya (collaboration agreement), CENTROGEO - SEDECULTA phrases (referenced in Agreement SEDECULTA-DASJ-149-04-2024) and [T'aantsil corpus project](https://taantsil.com.mx/info).
|
| 22 |
+
|
| 23 |
+
|
| 24 |
+
## Dependencies
|
| 25 |
+
|
| 26 |
+
Install gensim 4.0 version or greater.
|
| 27 |
+
|
| 28 |
+
```
|
| 29 |
+
$ pip install gensim
|
| 30 |
+
|
| 31 |
+
|
| 32 |
+
```
|
| 33 |
+
|
| 34 |
+
|
| 35 |
+
## Usage
|
| 36 |
+
|
| 37 |
+
See usage.py
|
| 38 |
+
|
| 39 |
+
```
|
| 40 |
+
import gensim
|
| 41 |
+
from gensim.models import Word2Vec
|
| 42 |
+
|
| 43 |
+
maya2vec = './model_512_60_5_-0.25_0.7308_3.35E-05'
|
| 44 |
+
|
| 45 |
+
# load global model
|
| 46 |
+
model = Word2Vec.load(maya2vec)
|
| 47 |
+
|
| 48 |
+
# Try out cosine similarity (dog, standing)
|
| 49 |
+
sim = model.wv.similarity("peek'", "waalak'")
|
| 50 |
+
print('''similarity("peek'", "waalak'")''', sim)
|
| 51 |
+
|
| 52 |
+
# Similarity between 'peek'' and 'waalak'': 0.9583
|
| 53 |
+
```
|
| 54 |
+
|
| 55 |
+
Cite the paper please: https:publishing_soon/paper
|
| 56 |
+
|
| 57 |
+
```
|
| 58 |
+
@article{maya2vec,
|
| 59 |
+
title={Generating Culturally and Linguistically Adapted Word Similarity Benchmarks for Indigenous Languages},
|
| 60 |
+
author={Molina-Villegas, Alejandro and Suro, Joel and Fernandez-Sabido, Silvia and Reyes-Magaña, Jorge},
|
| 61 |
+
journal={Inteligencia Artificial},
|
| 62 |
+
volume={29},
|
| 63 |
+
pages={103971},
|
| 64 |
+
year={2025},
|
| 65 |
+
publisher={IBERAMIA}
|
| 66 |
+
}
|
| 67 |
+
|
| 68 |
+
```
|
| 69 |
+
|
| 70 |
+
|
| 71 |
+
## License
|
| 72 |
+
|
| 73 |
+
Permission is hereby granted, free of charge, to any person obtaining a copy
|
| 74 |
+
of this software and associated documentation files (the "Software"), to deal
|
| 75 |
+
in the Software without restriction, including without limitation the rights
|
| 76 |
+
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
| 77 |
+
copies of the Software, and to permit persons to whom the Software is
|
| 78 |
+
furnished to do so, subject to the following conditions:
|
| 79 |
+
|
| 80 |
+
The above copyright notice and this permission notice shall be included in all
|
| 81 |
+
copies or substantial portions of the Software.
|
| 82 |
+
|
| 83 |
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
| 84 |
+
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
| 85 |
+
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
| 86 |
+
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
| 87 |
+
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
| 88 |
+
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
| 89 |
+
SOFTWARE.
|