1. 5


    Problem: HN has simply too much upvoting of mainstream news and megablog content. In this prototype I am assigning a technical score to all new articles for sorting purposes. All articles below a certain threshold get culled.

    1. 1

      How are you assigning the technical score?

      1. 1

        I always like to see how far the simplest approach takes me in the first iteration. I have three wordlists: CS terms, programming language keywords, and unix keywords. An article’s score is just a weighted sum of the # of words matched in each categories / # words in article. Current weights at time of writing = {‘cs’ : 1, ‘unix’ : 20, ‘code’ : 100}