adaamko/POTATO

Issues

• KeyError: "Constituency parser not trained with tag 'GW'"

  #63 opened 3 years ago by parthvishah
  0

• how hard to use this with a very large corpus?

  #42 opened 3 years ago by ym-han
  1

• Can this help with mining names (of people or concepts or groups or ...)?

  #44 opened 3 years ago by ym-han
  0

• Tests for Matchers

  #59 opened 3 years ago by adaamko
  2

• Building rules for homophobia detection - issues

  #54 opened 3 years ago by kuma-rtin
  0

• 'PotatoGraph' object has no attribute 'nodes'

  #50 opened 3 years ago by parthvishah
  2

• proper support for multi-label datasets

  #43 opened 3 years ago by recski
  0

• Instead of pickle, store Dataframes in TSV with penman formatted graphs

  #26 opened 3 years ago by adaamko
  0

• Resolving pip backtracking takes too long

  #39 opened 3 years ago by Eszti
  0

• Sometimes AMR to networkx conversion merges an edge into a nodename

  #41 opened 3 years ago by adaamko
  0

• error in training when using all classes of brise data

  #35 opened 3 years ago by recski
  2

• matching rule not highlighted in graphs when using regexes?

  #38 opened 3 years ago by recski
  0

• very good keyword not found by "suggest rules"

  #37 opened 3 years ago by recski
  0

• "suggest rules" should be able to run only on false negatives (i.e. only on what isn't covered already)

  #36 opened 3 years ago by recski
  0

• pattern matching bug (caused by self-loops?)

  #31 opened 3 years ago by recski
  1

• sentence split in multiple 4lang graphs not merged

  #32 opened 3 years ago by recski
  2

• The frontend should have an inference mode and better instructions

  #28 opened 3 years ago by adaamko
  0

• UI enhancement

  #20 opened 3 years ago by adaamko
  3

• Unsupervised

  #18 opened 3 years ago by GKingA
  0

• Save cache into language specific files

  #16 opened 3 years ago by Eszti
  0

• New scikit-criteria 0.5

  #23 opened 3 years ago by leliel12
  2

• Make usable for OneVsAll multi-label classification

  #17 opened 3 years ago by Eszti
  0

• Implement caching of the graphs for evaluation

  #21 opened 3 years ago by adaamko
  0

• Better examples in README for semeval

  #19 opened 3 years ago by adaamko
  0

• Ask twice to delete a rule

  #13 opened 3 years ago by adaamko
  0

• Make edges trainable

  #11 opened 3 years ago by adaamko
  0

• Unsupervised mode

  #9 opened 3 years ago by adaamko
  0

• Frontend UI improvement

  #6 opened 3 years ago by adaamko
  2

• Search for a proper programmatic generation of graphs, e.g. nx_pydot

  #8 opened 3 years ago by adaamko
  0

• Move *_to_graph functions to tuw_nlp repo

  #7 opened 3 years ago by adaamko
  0

• Frontend: option to select subset of rules to evaluate (default is that all rules are checked)

  #5 opened 3 years ago by adaamko
  0