News for hackers. Stuff that matters.

A comment on Show HN: Tracking voter discussion of the Presidential debates. Parent comment

Re:Show HN: Tracking voter discussion of the Presidential de... by jdunck on Wednesday October 03, 2012 @09:05PM (#4610444)
We have weeded out a bunch of stuff and are actually handling n-grams, but apparently not with enough weight since they aren't showing up ever.

What words do you think shouldn't be there?

[ Parent | on HN ]
Re:Show HN: Tracking voter discussion of the Presidential de... by endtime on Wednesday October 03, 2012 @10:18PM (#4610604)
Looks, Look, Good

Also, you should think about stemming - "Looks" and "Look" should be in an equivalence class.

[ Parent | on HN ]
Re:Show HN: Tracking voter discussion of the Presidential de... by erikrose on Wednesday October 03, 2012 @10:24PM (#4610617)
Perhaps we can bang a Snowball stemmer against it and get pretty good coverage on the whole language.
[ Parent | on HN ]