Graph displaying article topics

[dvmorozov]

Task

Implement main graph displaying topics of articles and their relations.

1. The graph should display set of topics (keywords) with relations. :heavy_check_mark:
2. Size of node should correspond to number of articles related to the topic. :heavy_check_mark:

3. 82.

4. Mouse moving over the node should display pop-up window showing name of topic and number of articles. :heavy_check_mark:
5. Edges of graph should connect topics related by articles. :heavy_check_mark: Only most important links are displayed.

6. 86.

7. 85.

Solution

1. Implement Python script extracting graph data from arxiv metadata. :heavy_check_mark: Related #75.
2. Filter-out data to show most important relations. :heavy_check_mark:
3. Implement graph page and JavaScripts to load and visualize data. :heavy_check_mark:
4. Use force-directed graph provided by d3.js. :heavy_check_mark:
5. Use GitHub as hosting platform. Use JavaScript as data format. No back-end code. :heavy_check_mark:
6. Add references to used components to the page. :heavy_check_mark:
7. Add hyperlink opening graph in separate browser tab. :heavy_check_mark:

Dependencies

https://pypi.org/project/ijson/

Data

https://www.kaggle.com/datasets/Cornell-University/arxiv https://arxiv.org/help/bulk_data

Commands

Extract downloaded data

gzip -d arxiv-public-datasets.gz

Tools

https://github.com/mattbierbaum/arxiv-public-datasets