Goodreads, but for research papers: meet PaperVote

From Letterboxd to Goodreads, a lot of us go to sites where others express their opinions to figure out what is worth our time based off our taste and needs. This is no different for researchers, where a common challenge is finding articles whose arguments are pertinent to their investigation. PaperVote is a new platform developed by an Imperial PhD student that provides exactly this: research paper recommendations based on people’s answer to the question: “Would you recommend this paper to someone else?” 

Now in public beta, PaperVote lets readers search by title, author, journal or DOI, drawing on sources including PubMed, Europe PMC, Crossref and DataCite, before linking through to the original paper record. After reading, users choose “recommend” or “would not recommend”. Users choose  their answer based off whether reading the paper was worth their time, not whether they agree with it’s conclusions. A paper stays “New/Unranked” until at least five distinct accounts have responded. The section “Community Choice” displays and ranks papers by calculating their recommendations minus non-recommendations. Accounts can be created without a name or email address, using a generated 20-digit recovery number that users must keep to regain access. Contributor tables recognise participation credits and the number of papers someone has responded to, but nobody’s recommendation carries extra weight. Appearing publicly is optional and requires a nickname. 

For Imperial students, a platform like PaperVote could prove useful when finding  reading beyond a course list, or choosing papers for a journal club. The rankings reflect reader opinion, not scientific validity, and do not replace peer review.