search

stonybrook-lin628-f19 / main

3 stars 2 forks source link

Potential Workshop Topics #2

Open thomas--graf opened 5 years ago

thomas--graf commented 5 years ago

I'll keep expanding this list throughout the semester as we come across interesting new challenges. If you are interested in tackling one of them, leave a note here.

CCG

  1. CCG analysis of an empirical phenomenon you've been working on
  2. Problematic cases for CCG (e.g. type differences between conjuncts); this will require some literature archeology and your own ingenuity
  3. Demo session of CCG parser(s), use of CCGbank
  4. Replicating an NLP paper (e.g. training and evaluating supertagging + A* parsing).
  5. Some NLP project of your own that uses CCG in some capacity

MGs

  1. MG analysis of an empirical phenomenon you've been working on
  2. Formal comparison of MG tree languages and TAG tree languages (hard!)
  3. Some project using the MG corpus

Subregular syntax

  1. Subregular analysis of an empirical phenomenon you've been working on
  2. Analysis of parasitic gaps
  3. Analysis of Coordinate Structure Constraint (CSC) and its ATB exceptions
  4. Handling exceptions to island constraints
  5. Case assignment (extending prior work by Vu.etal)
  6. Formal work on subregular tree transductions
  7. TSL extension for MGs that aren't in SMNF
  8. Subregular complexity of MGs with persistent features
  9. Does the SMNF-Island story work for a realistic grammar built from MGbank?
  10. Subregular complexity of adjunction

Misc

  1. Tutorial for some other formalism (e.g. MCTAG)
  2. Exploring the multi-shell VP analysis we used for maximally flexible adjunction (root + one extended projection for each argument).
vvulpes0 commented 5 years ago

I would like to investigate potential consequences of using DAGs in place of trees in syntax

alisalehi67 commented 5 years ago

Nazila + Ali Replicating an NLP paper (e.g. training and evaluating supertagging + A* parsing).

kkostyszyn commented 5 years ago

I'm interested in subregular complexity of MGs with persistent features

ghost commented 5 years ago

I am working on a project using the MG corpus

ay-la commented 5 years ago

I'll take "subregular complexity of TAG adjunction" for 400, Alex

joannechau commented 5 years ago

I am debating working on an extension of Case assignment using MG trees as inputs or working on CSC

pablolopezalonso commented 5 years ago

I'd like to do something with CCG parsers, either a demo session or an NLP project.

jiyanjingyu commented 5 years ago

I would like to apply CCG (or other theory) to analyze a syntax-prosody interface phenomenon.