Citations is a research tool used to visualize a research field as a DAG (Directed acyclic graph). This can be useful when trying to get familiar with a new research field.
The tool is in alpha version, so expect bugs and incompatabilities. Issues and contributions are highly welcome.
The tool uses Semantic Scholar API to get paper information, and creates a DAG out of the citations by recursively getting information on the cited papers. The papers are sorted from most cited (biggest) to least cited (smallest) and from oldest (blue) to newest (red). As default, the tool is showing only the most influential citations (according to Semantic Scholar's "highly influential" attribute) in order to make the graph more compact, but this filter can be disabled.
Some ideas on things to look at:
- The most impactful papers- look for the largest nodes in the graph.
- Newest papers- look for the red nodes.
- Trending papers- look for the largest red nodes.
Docker
Build the docker image
docker build . -t reference-graph
Let's make a graph starting from the paper "Meta clustering"
- We get the Semantic Scholar id from the url: 4127d83c9eef22918f16a7477ca6e60734c87d8b
- We use the graph.py script to create the graph from the paper's id, and scan 5 citation-levels deep:
docker run --rm -it -v $(pwd)/output:/myapp/output reference-graph --paper_id 2de0a40e9a5d4f1feb07d61af5a5d87a069653f0
For more arguments:
docker run --rm reference-grape --help