A flask-based prototype UI for annotation of qualia relations between FrameNet Lexical Units inferred from an external knowledge base

Clone the repository cd to it.

git clone https://github.com/slowwavesleep/QualiaAnnotationUI.git

cd QualiaAnnotationUI

Create and activate a virtual environment for the project. I suggest using Anaconda.

conda env -n qualia python=3.7

conda activate qualia

Install the requirements.

pip install -r requirements.txt

Run the application.

flask run

To annotate a different dataset replace qualia_relations.csv in resources directory with a new file. Or alternatively, change data_path in config.yml to point to the new file. In both cases the file must in CSV format (comma-separated specifically) with the following columns:

• index
• id_lu_1
• word_1
• pos_1
• fnDefinition_1
• relation
• id_lu_2
• word_2
• pos_2

The tool uses SQLite to persist user annotations between the launches. Made annotations may be downloaded in CSV format.