Network of Estonian Translated Literature

Author: Krister Kruusmaa

Institutions: National Library of Estonia, Tallinn University

License: MIT

Repository for graph-based analysis and interactive vizualiation of Estonian translated literature, connecting thousands of authors and translators via published works.

Related publications:

• Maris Saagpakk, Daniele Monticelli, Krister Kruusmaa (2024). Eesti tõlkeajaloo digigalaktika toob ilmsiks kultuuri ja ühiskonna aluskoe. ERR Novaator 01.04.2024. https://novaator.err.ee/1609299408/eesti-tolkeajaloo-digigalaktika-toob-ilmsiks-kultuuri-ja-uhiskonna-aluskoe
• [TBA]

Interactive app: https://data.digar.ee/tkirjandus/index.html

Overview and usage

The project contains a back-end and a front-end module. The network/ directory contains scripts to generate a graph object from curated metadata of the Estonian National Bibliography, as well as analysis notebooks and plots. The app directory contains code for an Typescript-based interactive web app using Graphology and Sigma modules. To build the network locally, follow these steps:

1. Install the required Python modules via pip using the reqiurements.txt file.
2. Download the Curated ENB dataset and place it into the network/data/raw/ directory.
3. Run ./network/src/create_graph.py. This creates two files - a JSON file in the ./network/data directory and a GEXF file in the ./network/data/gephi directory. For computation analyses, use the JSON file.
4. Open the GEXF file with Gephi to to adjust layout and node size, then export it via File -> Export -> Graph object (Gephi v. 0.10.1). Use the same file name as for the existing .gexf file (default data).
5. Run ./network/src/update_data.py. This updates the JSON file with the custom layout from Gephi, creating an updated data file in network_data.
6. Move the updated data file to the ./app/public/ directory.
7. Use npm run build to build the app (npm must be installed).
8. Run an HTTP server in the ./app/build/ folder.