Open Journal Matcher: open journal discovery for everyone
summary
A web app that matches a draft abstract to open access journals is live and working: it compares a user abstract against about 5600 DOAJ journals (limited to ~100 recent abstracts per title), scores similarity using spacy word vectors, and performs thousands of parallel comparisons via Google Cloud Functions with a Flask front end. Results return the top five recommendations; accuracy is reasonable, multilingual coverage is limited, and ethical and metadata-quality caveats apply.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:08Overview
- 3:05Other journal recommenders
- 4:14Directory of Open Access Journals
- 5:00ISNs
- 6:39Flask
- 9:18Under the hood
- 9:50Spacey
- 10:41Word vectors
- 16:12Ethical incompleteness
related talks
37mA Shiny application for job matching
27mStardust: Digital Tooling for Earthly Science & Education
24mOpenDS4All: An open education project to accelerate the the development of data science curricula
35mProduct fairness testing for Machine Learning developers
01h 05mThe Linked Data Revolution is Here, Are you Ready?
40mData Doesn't Lie: How to Influence Open Source Programming Through..
14mInformal Learning Communities: The Other Massive Open Online 'C'
15mF8 Refresh | Supporting academic research on Facebook