This repo contains the software that was used to conduct the experiments reported in our article titled "Improving Named Entity Recognition by Jointly Learning to Disambiguate Morphological Tags" [1] to be presented at COLING 2018.
We recommend using the helper scripts for conducting experiments. The scripts named helper-script-*
run the experiments in the paper with given hyper parameters. Beware that the option -j6 will make the script run
six experiment configurations at once.
bash ./scripts/default/helper-script-to-run-the-experiment-set-small-sizes.sh campaing_name | parallel -j6
For the reporting part to work, you should set up a working sacred
environment, which is very easy if you choose a filesystem based storage. You can find an
example of this in the helper script found in ./scripts/TRUBA folder.
This project do not have a designated tagger script for now but you can obtain the output in eval_dir.
You should provide the text in tokenized form in CoNLL format. The script will tag both the development and
testing files and produce files in ./evaluation/temp/eval_logs/. If you need this and want to contribute by coding
and sharing it with the project, you are welcome.
To reproduce the experiments reported with our model, you can use Docker
and build a replica of our experimentation environment.
To build:
docker build -t yourimagename:yourversion .To run:
docker run -ti -v `pwd`/dataset:/opt/ner-tagger-dynet/dataset -v `pwd`/models:/opt/ner-tagger-dynet/models yourimagename:yourversion python train.py --train dataset/gungor.ner.train.small --dev dataset/gungor.ner.dev.small --test dataset/gungor.ner.test.small --word_dim 300 --word_lstm_dim 200 --word_bidirect 1 --cap_dim 100 --crf 1 --lr_method=adam --maximum-epochs 50 --char_dim 200 --char_lstm_dim 200 --char_bidirect 1 --overwrite-mappings 1 --batch-size 1You should create or set permissions accordingly for `pwd`/dataset and `pwd`/models.
[1] Gungor, O., Uskudarli, S., Gungor, T., Improving Named Entity Recognition by Jointly Learning to Disambiguate Morphological Tags, 2018, COLING 2018, 19-25 August, (to appear).
Dockerfile should reside in the root directory. What can we do about this? Maybe the DIP should be informed about this
kind of Dockerfiles in some way?
I will provide the model files as a single tar file.
docker build -t yourimagename:yourversion .docker run -ti -p 8080:8080 -v `pwd`/dataset:/opt/ner-tagger-dynet/dataset -v `pwd`/models:/opt/ner-tagger-dynet/models yourimagename:yourversion python main.py --command webapp --model_path model-00002714 --model_epoch_path model-epoch-00000026 --port 8080