AcademiaChain: Your Erdős Number, Generalized
I shipped a new side project yesterday: AcademiaChain 🌌 — find the shortest co-authorship path between any two scholars, your Erdős number, generalized.
💡 Where this came from
Back in college, I stumbled on a graph theory problem: in any large enough group of people, you can always find a subset who are either all mutually acquainted or all mutual strangers. It sparked an idea — if every human is a node and “knowing each other” is an edge, the whole world is one giant connected graph. I pitched it to a professor back then. They said, “it’s possible.” The idea faded… until recently.
I still wanted to build it, but without touching private or restricted social data (Facebook-style graphs are off limits for good reason). Then it hit me: public academic metadata. If two scholars co-authored a paper, that’s an edge — openly available, no privacy trade-off.
🌐 What it does
Type in two researchers’ names, and AcademiaChain traverses the OpenAlex open scholarly graph — live, with no pre-built database — to find the chain of co-authors connecting them, along with the bridging papers (title, year, DOI). Progress streams to your browser in real time, and a match renders as a glowing constellation graph: drag the stars, and the bonds between scholars grow thicker and brighter the more papers they share. Find a path, and you can generate a shareable Academic Lineage Certificate poster.
Under the hood it’s a bidirectional breadth-first search — two searches growing from each name toward the middle — which turns what would be a 10-second timeout into a sub-second collision, with pruning to skip giant consortium papers and keep the search fast and relevant.
😂 The humbling benchmark
I tested it on my own academic network against Terence Tao (Fields Medalist). Result: No Path Found within 4 degrees. Maybe it’s the strict disciplinary isolation of modern research — or maybe my pruning parameters just weren’t greedy enough. Either way, it was a good reminder that “small world” claims are easier to state than to prove.
⚠️ A quick heads-up on data quirks
OpenAlex’s autocomplete applies fairly aggressive “popularity truncation,” so younger scholars or common names can occasionally get buried in the quick dropdown. I added an Advanced Search Modal that lets you filter by institution keyword for precise identity locking — but edge cases exist. It’s a work in progress.
🔒 Privacy, by design
Everything comes from OpenAlex’s public, published academic metadata — no private data, no database, no logs. Any API key or academic email you configure lives only in your browser’s local storage.
🚀 Try it
- Live demo: academiachain.netlify.app
- Source (100% open): github.com/jeromecy/AcademiaChain
Try it with your own name, test the boundaries, or dig into the bidirectional BFS implementation — and let me know how small (or not) your corner of academia turns out to be.