MT-ComparEval is a tool for comparison and evaluation of machine translations. It allows users to compare translations according to several criteria, such as:
MT-ComparEval also plots a chart with absolute differences of metrics computed on single sentences and a chart with values obtained from paired bootstrap resampling. MT-ComparEval is distributed under Apache 2.0 license with an exception of Highcharts.js library which is distributed under Creative Commons Attribution-NonCommercial 3.0 License.
When using MT-ComparEval please cite:
For a user-focused show-case study explaining most of the features, see:
The default configuration shipped with the software includes only the basic BLEU-related metrics. If you are interested only for these metrics, proceed to the next chapter "Basic Installation".
There is the possibility to enable additional metrics. At the moment, the tool "Hjerson" is supported but it is optional. Its entries are commented out in the file config.neon , in folder app\config, after downloading the software. If you are interested in enabling Hjerson, you should edit this file and uncomment (remove the leading # from) the lines that contain the name of the metric. Then proceed to the installation as following.
MT-ComparEval has been designed and tested on Ubuntu and Debian systems. Such an operating system is therefore suggested.
In order to be able to run MT-ComparEval several dependencies have to be installed. On Ubuntu 20.04, these dependencies can be installed with the following commands:
sudo apt install sqlite3 php7.4-cli php7.4-sqlite3 curl php7.4-mbstring php7.4-xml php7.4-curlOn Ubuntu 18.04 use:
sudo apt install sqlite3 php7.2-cli php7.2-sqlite3 curl php7.2-mbstring php7.2-xml php7.2-curlOn Ubuntu 16.04 use:
sudo apt install sqlite3 php7.0-cli php7.0-sqlite3 curl php7.0-mbstring php7.0-xmlOn Ubuntu 14.04 use:
sudo apt-get install sqlite3 php5-cli php5-sqlite curlDownload or clone the application and set your commandline within the new directory. Then the application can be installed with the following command:
bash bin/install.shDuring the installation you may be asked to enter GitHub OAuth token. Just follow the instructions (open the url in your browser, generate the token and enter it).
C:\Program Files\ but it may vary depending on the language and the installation of your system.sqlite. Download SQLite Precompiled Binaries for Windows (probably 64bit, this depends on your version of Windows) and the SQLite Tools Bundle. Extract the contents directly in the subfolder sqlite that you created previously, withouth creating or including any other subfolder. A permissions question will appear, this is normal.C:\ and create a directory tools. Download MT-Compareval, open the zip file and copy the folder MT-ComparEval-master into C:\tools directory. Rename the folder from MT-ComparEval-master to MT-ComparEval.C:\tools\MT-ComparEval\bin\win and double click on install or install.bat. This will perform the installation. During the installation you may be asked to enter GitHub OAuth token. Just follow the instructions (open the url in your browser, generate the token and enter it).To start MT-ComparEval two processes have to be run:
bin/server.sh which starts the application server on the address localhost:8080
(you can can check/adapt app/config/config.neon first to set the main title, set of metrics etc. See the default config.)bin/watcher.sh which monitors folder data for new experiments and tasks (the data folder must exist before you run bin/watcher.sh.)server or server.bat to start the web server. Leave the window openwatcher or watcher.bat to start the watcher. Leave this window open toodata folderFolder data contains folders with experiments (e.g. EN-CS-WMT15), which contains subfolders with tasks for each experiment (e.g. MOSES). For example:
data/
├─ EN-CS-WMT15/
│ ├─ source.txt
│ ├─ reference.txt
│ ├─ CHIMERA/
│ │ └─ translation.txt
│ └─ NEURAL-MT/
│ └─ translation.txt
└─ EN-DE-WMT15/
├─ source.txt
├─ reference.txt
├─ MOSES/
│ └─ translation.txt
└─ NEURAL-MT/
└─ translation.txtEach folder corresponds to one experiment and it should contain the following files:
source.txt - a plain text file with sentences in source language (one sentence per line).reference.txt - a plain text file with reference translations (in target language).config.neon - (optionally) a configuration file with the following structure:
name: Name of the experiment
description: "Description of the experiment\n can be multiline"
source: source.txt
reference: reference.txtSee http://ne-on.org/ for the syntax of neon files.
The source and reference needs to be defined only if you you choose non-default file names (not source.txt and reference.txt).
Individual machine translations called tasks are then stored in subfolders with the following files:
translation.txt - a plain text file with translated sentencesconfig.neon - (optionally) a configuration file with the following structure:
name: Name of the task
description: Description of the task
translation: translation.txt
precompute_ngrams: trueA curl command to create an experiment:
curl -X POST -F "name=experiment name" -F "description=description" -F "source=@source.txt" -F "reference=@reference.txt" http://localhost:8080/api/experiments/uploadA curl command to create a task:
curl -X POST -F "name=task name" -F "description=description" -F "experiment_id=1" -F "translation=@translation.txt" http://localhost:8080/api/tasks/uploadFor deleting experiments via API use api/experiments/delete/<id>.
data/.../ (or if you want to reimport the task after watcher is restarted, delete the hidden files .imported and .notimported)sqlite3 storage/database "SELECT id, name FROM tasks WHERE experiments_id=XYZ";sqlite3 storage/database "sqlite3 storage/database "DELETE FROM tasks WHERE id=ABC";"