News for hackers. Stuff that matters.

A comment on Show HN: News headline analysis of over 140,000 headlines over several years.

Re:Show HN: News headline analysis of over 140,000 headlines... by chris_va on Wednesday October 23, 2013 @06:52PM (#6602310)
If you want to have a fun time with this, break it down by title keywords (e.g. politician names). You can get a very good break down of political bias by news source.

In a more automated way, you can extract all keywords from your titles, and then auto-extract the top keywords that are the most dividing.

(Source: former Google News TL, we used to have way too much fun doing stuff like this)

[ Parent | on HN ]
Re:Show HN: News headline analysis of over 140,000 headlines... by Luiz7 on Wednesday October 23, 2013 @07:11PM (#6602416)
Yeah we talked about that as being the obvious next step but we had neither the time nor the skillset off the bat for it. A noSql DB would probably be the better option for that as well. (correct me if i'm wrong)
[ Parent | on HN ]
Re:Show HN: News headline analysis of over 140,000 headlines... by chris_va on Wednesday October 23, 2013 @08:26PM (#6602701)
With 150K titles, you don't have enough data to warrant a no-sql database. You can set up mysql or postgres in about 10 minutes and dump in the data.
[ Parent | on HN ]