It’s just simple keyword matching in the order of appearance on the list. I did something similar with the DMOZ directory: https://www.locserendipity.com/DMOZ.html

Also, as an experiment, I developed an auto-summarizer, scraped and summarized all of the edu domains on DMOZ, pared the non-active ones, and the ordered them by the amount of content (more content ranks higher). The result is this search engine: https://www.locserendipity.com/edu.html