I need an English parser written in Java with a small memory/processing input. I have used the Stanford parser, but it is rather heavy. I am also considering using the link grammar parser, which is fast and small, but would require me to write several more grammar rules for my domain. The goal is to extract a semantic structure from an utterance, so anything that can lead easily to that would be nice. Can anyone suggest something good to use?

  • @NateGlenn: I think the right answer might depend on your needs. Do you need dependencies, or bracketing? Do you only need chunk structure, or do you need hierarchical structure? – jogloran Aug 3 '12 at 5:51
  • @jogloran I actually may have misformulated the question. I do need structure, but I plan on extracting semantic information from it. So anything that can lead to a semantic structure is fine. The LG parser just emits links, which I think would be easy enough to use. – Nate Glenn Aug 3 '12 at 13:48
  • Any particular reason the it should not use as much memory as the Stanford parser? If you are parsing sentences one by one, you could even create a server for the parser and send queries over a socket. You only need to load the parser once. It takes about 2 seconds per sentence after that. – prash Aug 3 '12 at 15:23
  • I suggested the same thing to my boss; but for this project, the entire thing has to run on an Android tablet, even with no internet connection. – Nate Glenn Aug 3 '12 at 15:50
  • @NateGlenn If you need to make clarifications on the question but you don't know how to make it, please consider using our Linguistics Chat or posting a question in Linguistics Meta. :) Link the discussion here in case you use Meta or if you create a new room in Chat. – Alenanno Aug 3 '12 at 16:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.