wiki:questionnaire

1. what approach your system can be categorized into?

  • hand-coded rules
  • statistical (e.g. machine learning)
  • hybrid
  • other (free description)

2. fundamental approach of your system (you can choose more than one)

  • overlap
  • alignment
  • first-order, higher-order logic
  • transformation
  • other (free description)

3. what kind of information does your system use? (you can choose more than one)

  • character, word overlap
  • syntactic structure
  • predicate-argument structure
  • named entity information
  • relation between entities
  • relation between events
  • temporal information
  • numerical information
  • other (free description)

4. what kind of resources does your system use? (you can choose more than one)

  • synonym, antonym
  • hypernym, hyponym
  • meronym, holonym
  • paraphrase/entailment rules
  • relations between events (e.g. cause-effect)
  • other (free description)

5. please enumerate the linguistic analyzers and the resources your system used

  • e.g. Japanese WordNet, Wikipedia, Google, etc…
  • (free description)

6. is there any manual intervention in making the results of the formal run?

  • yes
    • (if yes) please describe the detail of the manual intervention
  • no

7. what information provided by the organizers does your system use? (JA only)

  • unidic morph
  • unidic syntactic
  • knp morph
  • knp syntactic
  • knp predicate-argument-structures
  • TSUBAKI search results
  • none
  • other (free description)