HyperStorylines
Associated publication
Demo
A running version of the tool can be found here.
Installing with Docker
Requirements
- Python 3
- Docker 23.0.3
- docker-compose 1.29.2
Build and run the Docker image
sudo docker-compose up -d --build
sudo docker-compose run web python manage.py index_default_datasets
Create new datasets
By default, Nested Storylines comes with 3 datasets:
- A subset of the DBLP database with articles from IEEEVIS, EUROVIS and CHI.
- An anonymized news articles dataset, based on the work with data journalists described in the paper
- A co-authorship graph of ILDA, AVIZ and people from other teams. This dataset doesn't have text to index so text search queries by the name of the entities.
To create a new dataset, check out the dataset_generation folder.
It contains necessary libraries to create a hypergraph (lib/hypergraph.py) that can be
exported to Nested Storylines to read it (hypergraph_instance.export_to_json(filename)).
It also contains some compatibility methods to export them in the PAOH vis format.
For the DBLP sample dataset and the news articles dataset, there is a folder in the
dataset_generation folder with example scripts on how to create them.