Tsunnami commited on
Commit
617be3b
·
verified ·
1 Parent(s): 1bfc82e

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +12 -19
README.md CHANGED
@@ -7,28 +7,21 @@ sdk: static
7
  pinned: false
8
  ---
9
 
10
- # MEET-MR (Machine Translation English–Thai MQM and Ranking)
 
11
 
12
- Welcome to the official Hugging Face organization for the **MEET-MR** project. Our mission is to advance the state-of-the-art in machine translation evaluation, specifically focusing on low-resource and typologically distant language pairs like **English–Thai**.
13
 
14
- ## 🚀 Our Mission
15
- Evaluation remains a critical bottleneck in MT research. Automatic metrics often fail to capture the nuances of adequacy and fluency in Thai. We bridge this gap by providing high-quality, human-annotated datasets and fine-tuned quality estimation models that align closely with human judgment.
 
16
 
17
- ## 📊 Key Contributions
18
- - **MEET-MR Dataset**: A comprehensive benchmark containing over 20,000 segments across 9 domains, featuring professional MQM error annotations and 10-way human preference rankings.
19
- - **Evaluation Models**: Fine-tuned COMET-based metrics (COMET-22 and COMET-kiwi) specifically optimized for English-Thai translation quality estimation.
20
- - **Linguistic Diversity**: Data curated from diverse sources including medical terminology, legal treaties, technical manuals, and conversational text.
21
-
22
- ## 🏫 Affiliations
23
- This project is a collaborative effort between:
24
- - **Department of Computer Engineering**, Faculty of Engineering, Chulalongkorn University.
25
- - **Department of Mathematics**, King’s College London.
26
-
27
- ## 📖 Publications
28
- Our work was introduced at **EACL 2026**.
29
- > Phuangrot, P., et al. (2026). *Machine Translation Evaluation English-Thai MQM Ranking Dataset*. Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics.
30
-
31
- **[View Paper on OpenReview](https://openreview.net/forum?id=0IbK0VeN8y)**
32
 
 
 
33
  ---
34
  *For inquiries or collaborations, please reach out via our primary repository issues or the contacts listed in our paper.*
 
7
  pinned: false
8
  ---
9
 
10
+ # MEET-MR
11
+ **Advancing English–Thai Machine Translation Evaluation**
12
 
13
+ MEET-MR provides high-quality human-annotated benchmarks and models to bridge the evaluation gap for the English–Thai language pair. Our work focuses on fine-grained error analysis (MQM) and human preference alignment.
14
 
15
+ ### 🛠 Key Resources
16
+ * **[MEET-MR Dataset](https://huggingface.co/datasets/MEET-MR/MEET-MR):** 20k+ segments across 9 domains with MQM labels and 10-way human rankings.
17
+ * **[Evaluation Models](https://huggingface.co/MEET-MR/COMET-21-MEET-MR):** Fine-tuned COMET metrics optimized for English-Thai quality estimation.
18
 
19
+ ### 📖 Publication
20
+ Introduced at **EACL 2026**:
21
+ *"Machine Translation Evaluation English-Thai MQM Ranking Dataset"*
22
+ [Read the Paper](https://openreview.net/forum?id=0IbK0VeN8y)
 
 
 
 
 
 
 
 
 
 
 
23
 
24
+ ### 🏫 Affiliations
25
+ * Department of Computer Engineering, **Chulalongkorn University**
26
  ---
27
  *For inquiries or collaborations, please reach out via our primary repository issues or the contacts listed in our paper.*