learn.ifs.name

Coreference Resolution and Entity Linking

Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 24

6 cards in “24.5 Classifiers using hand-built features” · whole deck

24.5 Classifiers using hand-built features

  1. Why are feature-based coreference classifiers still useful?
    They are lightweight when compute or data are scarce, and their features help error analysis even for neural systems.
  2. What feature classes do feature-based coreference classifiers use?
    Features of the anaphor, of the candidate antecedent, and of the pair; entity-based models add features of the antecedent’s cluster and of the relation between the anaphor and the cluster’s mentions.
  3. Give example features for the pair (Victoria Chen, she).
    head wordsVictoria / she
    attributesSg-F-A-3-PER for both
    length2 / 1
    mention typeP / Pr
    entity shapeP-Pr-D
    sentence distance1
    i-within-iF
    exact string / head matchF
  4. Which hand-built features are particularly useful for coreference?
    Exact string match, entity headword agreement, mention distance; for pronouns exact attribute match and i-within-i; for nominals and names word inclusion and cosine.
  5. Why are feature conjunctions important in feature-based coreference?
    Moving from individual features to conjunctions gained about 4 F1 points; conjunctions can be hand-designed, all pairs, or learned by decision trees or random forests.
  6. Which hand-built features still help neural coreference models?
    Not shallow ones like string match or mention type (contextual embeddings cover them), but mention length, distance between mentions, and genre.

Source