A Unified Generative Framework for Various NER Subtasks

This paper adopts the text-to-text format from Google’s T5 to NER task and reports somewhat competitive results to biaffine-ner. They employ BART-Large to generate the indices and tags of named entities in the input sentence.


  • T5 should be emphasized as the source where this text-to-text formulation is proposed and where they got inspiration.
  • Is there a better way to interpolate the pointer/tag distribution with the token distribution learnt by the pre-trained decoder?
  • Using seq2seq for NER seems to be an overkill as it’s very slow, as reported by the authors in their appendix.
  • Regarding “the sentencepiece tokenization used in T5 will cause different tokenizations for the same token, making it hard to generate pointer indexes to
    conduct the entity extraction”, not sure why this is a problem. You can still feed pre-tokenized tokens into sentence piece and get a deterministic tokenization for each token.
  • 5: Transformative: This paper is likely to change our field. It should be considered for a best paper award.
  • 4.5: Exciting: It changed my thinking on this topic. I would fight for it to be accepted.
  • 4: Strong: I learned a lot from it. I would like to see it accepted.
  • 3.5: Leaning positive: It can be accepted more or less in its current form. However, the work it describes is not particularly exciting and/or inspiring, so it will not be a big loss if people don’t see it in this conference.
  • 3: Ambivalent: It has merits (e.g., it reports state-of-the-art results, the idea is nice), but there are key weaknesses (e.g., I didn’t learn much from it, evaluation is not convincing, it describes incremental work). I believe it can significantly benefit from another round of revision, but I won’t object to accepting it if my co-reviewers are willing to champion it.
  • 2.5: Leaning negative: I am leaning towards rejection, but I can be persuaded if my co-reviewers think otherwise.
  • 2: Mediocre: I would rather not see it in the conference.
  • 1.5: Weak: I am pretty confident that it should be rejected.
  • 1: Poor: I would fight to have it rejected.

0 投票者