2

I am trying to use the algorithm in Qiu et al., 2011, "Opinion Word Expansion and Target Extraction through Double Propagation" for my research. However, in the paper, they use Minipar for dependency parsing and I would prefer to use Stanford Parser.
I found the description of the dependencies from Minipar but I find them very vague. For Stanford Parser, I am referring to the list here.

Since I am not a linguist, I am having trouble finding the mapping myself. The list of the dependencies that I need to map and my best guesses for the matching dependencies is the following:

  • mod: probably amod, also nummod?
  • pnmod: possibly nmod, also nmod:npmod? nmod:poss? nmod:tmod?
  • subj: probably nsubj, could also include nsubjpass, csubj and csubjpass
  • s: no idea
  • obj: probably dobj
  • obj2: iobj?
  • desc: Not even mentioned in the GATE list

Thanks for the help!

0

I found the paper "Improving Opinion Aspect Extraction Using Semantic Similarity and Aspect Associations", Liu et al., 2016, where the authors use the algorithm by Qiu but with Stanford Dependencies, i.e., exactly how I want to do it. They use the dependencies amod, prep, nsubj, csubj, xsubj, dobj and iobj (and conj). I don't know the exact mapping (if it even exists) but it isn't important for my work.

| improve this answer | |

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.