0

I'm looking for a list of English part of speech sequencing rules (e.g. "a determiner cannot be followed by a verb"). Thought it would be easy but I couldn't find an actual list of more than several examples. Any ideas?

Thanks.

user1049052
  • 123
  • 1
  • 4
  • 10

1 Answers1

1

The problem of make a "list of POS constraints" lies in the fact that those constrants will mainly depends on discourse domain.

I think you can face it from a n-gram approach. You can make POS tagging over a specific corpus (wikipedia articles for certain topic for example) then generate 2-grams or 3-grams (using grams of words) and calculate their frequencies, so you will get the most/less frequent POS combinantions. Finally, you can think about those POS combinations which not even appeared in the frecuency list, such sequences may be called "part of speech constraints".

Jason Angel
  • 2,233
  • 1
  • 14
  • 14