1. Currently, the .json files in the directory _data_ are manually fetched; so if new files are added to that directory, the app code must be changed to handle the new files
2. Since the graph is a multi-edge directed graph, the visualization tool uses a setting called _smooth edges_ to differentiate between multiple edges between the same two nodes
* With large networks, however, this takes a longer time to render the graph (see next sub-section)
* With large networks, however, this takes a longer time to render the graph
## Example: Network of _Friedrich III._ (925 nodes, 4468 edges in graph_v02.json)
# Graph files
1. graph_v10.json: As described in paper
2. graph_v11.json:
* adds nobility titles and named entity types to each node
* [bugfix] issuers and normal named entities are resolved separately ("i: Alfons" and "e: Alfons")
* [bugifx] named entites mistakenly ending on "v." (e.g. "Dieter v.") get extended by one word ("Dieter v. Minzenberg")
* [bugfix] named entities with <= 2 chars get removed
### macOS, Firefox:
* with smooth edges set to _true_ it takes 3:09 min. until rendered
* with smooth edges set to _false_ it takes 0:52 min. until rendered
# Reference
If you make use of the contents of this repository, please cite the following paper:
### macOS, Safari:
* with smooth edges set to _true_ it takes 1:06 min. until rendered
* with smooth edges set to _false_ it takes 0:09 min. until rendered
However, currently we default to smooth edges _true_, so there are two ways to handle larger networks:
1. Collapse multiple edges into one in those networks (and show details only when an edge is clicked on)
2. Refine the graph so that it contains fewer edges
```
@inproceedings{opitz2018ri,
title={{Induction of a Large-Scale Knowledge Graph from the Regesta Imperii}},
author={Opitz, Juri and Born, Leo and Nastase, Vivi},
booktitle={2nd Joint SIGHUM Workshop on Computational Linguistics for Cultural Heritage, Social Sciences, Humanities and Literature},
