Comments on: Grammar Checkers – Not so much magic

By: rsmudge

rsmudge — Wed, 09 Dec 2009 15:15:51 +0000

Welcome, always glad to have readers. Writing stuff is more fun that way.

By: flashrash

flashrash — Wed, 09 Dec 2009 14:54:39 +0000

you just found a new daily reader

By: rsmudge

rsmudge — Wed, 18 Nov 2009 10:14:31 +0000

I think it’s feasible to do some rule mining by developing structures that capture patterns of language use from a good data set and then finding the difference between this structure and one taken from a set of mixed good and bad data. I probably won’t be the one to look further into this though. The approach AtD takes has limitations and they limit what I can catch but it works well enough that I haven’t been drawn to investigate this more.

By: sanjeev sharma

sanjeev sharma — Wed, 18 Nov 2009 04:54:58 +0000

hi
i agree with this that we should think in a different way. now the researches are just to accelerate the pre developed grammar checker but i think we should think about the new techniques that can be used to improvr the grammat checker.

By: rsmudge

rsmudge — Tue, 01 Sep 2009 03:42:55 +0000

Hah, thanks for the comment love Bishop Song. I’ll work on injecting more of my personality into this blog once I get the past years worth of NLP self-study off my chest and dumped into this blog. Something in there has to be new

By: Dug Song

Dug Song — Tue, 01 Sep 2009 02:35:28 +0000

I like that the possibly related post is “Monkeys Sense Poor Grammar”… who needs parse trees, part-of-speech tagging, and all that NLP mumbo-jumbo when you can put a room full of trained monkeys to work? Automation != computerization