README.md

metadata

license: apache-2.0
language:
  - en
  - de
  - ru
  - zh
tags:
  - mt-evaluation
  - WMT
size_categories:
  - 100K<n<1M

Dataset Summary

This dataset contains all MQM human annotations from previous WMT Metrics shared tasks and the MQM annotations from Experts, Errors, and Context.

The data is organised into 8 columns:

• lp: language pair
• src: input text
• mt: translation
• ref: reference translation
• score: MQM score
• system: MT Engine that produced the translation
• annotators: number of annotators
• domain: domain of the input text (e.g. news)
• year: collection year

You can also find the original data here. We recommend using the original repo if you are interested in annotation spans and not just the final score.

Python usage:

from datasets import load_dataset
dataset = load_dataset("RicardoRei/wmt-mqm-human-evaluation", split="train")

There is no standard train/test split for this dataset but you can easily split it according to year, language pair or domain. E.g. :

# split by year
data = dataset.filter(lambda example: example["year"] == 2022)

# split by LP
data = dataset.filter(lambda example: example["lp"] == "en-de")

# split by domain
data = dataset.filter(lambda example: example["domain"] == "ted")

Citation Information

If you use this data please cite the following works:

• Experts, Errors, and Context: A Large-Scale Study of Human Evaluation for Machine Translation
• Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain
• Results of WMT22 Metrics Shared Task: Stop Using BLEU – Neural Metrics Are Better and More Robust