A Custom Jupyter Widget for selecting part of text and assingning it one of tags.
Use the widget with the active learning tool:
https://github.com/IINemo/active_learning_toolbox/tree/seq
An example of use case can be found in workspace/example.ipynb
Note: if you have less than 10 tags you don't need to pass colors, otherwise you have too.
The result is represented via widgets public field res.
It can be in 1 of 3 states:
- Empty: nobody selected anything in widget
- Containing string of
None: text doesn't contain any squences and user checked this widget as "Done" - Containing dicts with selected sequences
To install this extension simply run folowing comands:
$ pip install text_selector
$ jupyter nbextension enable --py --sys-prefix text_selector
To install for jupyterlab
$ jupyter labextension install text_selector
For a development installation (requires npm) run in terminal:
$ git clone https://github.com/randomunrandom/text_selector.git
$ cd text_selector
$ pip install -e .
$ jupyter nbextension install --py --symlink --sys-prefix text_selector
$ jupyter nbextension enable --py --sys-prefix text_selector
$ jupyter labextension install js
To create and start docker container run:
$ git clone https://github.com/randomunrandom/text_selector.git
$ cd text_selector
$ docker build -t text_selector .
$ docker run -p 8888:8888 -it --rm --name text_selector text_selector
if port 8888 is occupied change ${port} to any available port
$ docker run -p ${port}:8888 -it --rm --name text_selector text_selector
@inproceedings{shelmanov2019bibm,
title={Active Learning with Deep Pre-trained Models for Sequence Tagging of Clinical and Biomedical Texts},
author={Artem Shelmanov and Vadim Liventsev and Danil Kireev and Nikita Khromov and Alexander Panchenko and Irina Fedulova and Dmitry V. Dylov},
booktitle={Proceedings of International Conference on Bioinformatics & Biomedicine (BIBM)},
year={2019}
}