Citegraph is and will always be free and open-source without ads. If you'd like to build and host the platform yourself, please follow the guideline on GitHub.

This website was initially built as a demo of JanusGraph, an open-source graph database, but we felt it could be useful for researchers, so we are committed to developing and maintaining it. We don't accept donation but contribution and feedback are always welcome.

The information in Citegraph comes from open-source datasets. Sometimes, there might be mistakes or gaps in this data. One common issue is confusion between authors who share the same name. This can lead to an author having more than one page, or different authors being mixed up as one person.

As of now, the data set purely comes from DBLP-Citation-network V14, an open-source citation dataset published on 2023-01-31.

If you spot any mistakes or missing data, please let us know. You can do this by raising an issue on our GitHub page. We'll do our best to fix it manually.

Citegraph draws an edge from author A to author B if A has ever written a paper X that cited a paper Y written by author B, i.e. A ---writes--> X ---cites--> Y <--writes--- B. We call this edge as `REFER` edge. The edge also has a counter which records how many times A has `referred` B.

`Referers` of an author X are people who have `referred` X. If you are looking for potential collaborators or people who can write you reference letters, this feature can be handy.

`Referees` of an author X are people who X has `referred`. If you are interested in researcher X's works, this feature can help you find out whose works X has been following.

Pagerank is calculated using the classic page rank algorithm by PageRankVertexProgram. All papers are involved in the calculation and they vote for the papers they cite until the pagerank score converges. We scale the score values so that the average pagerank score is 1.0. That means, if a paper's pagerank score is larger than 1, then it has more impacts than an average paper in Citegraph.

The pagerank of an author is simply a summation of all their papers' pageranks. If an author's pagerank score is larger than the number of their papers, then this author's impact per paper is higher than an average author.

We visualize pageranks in the visualizer. A larger circle means that paper or author has a higher pagerank and vice versa.

Note: the author order of a paper is not taken into consideration when computing authors' pageranks. Also, we assume one citation represents one unweighted positive vote, which may not always be the case.