FANDOM


To parse natural text to logical formulas (different from frame-semantic parsing or dependency-based semantic parsing).

Problem: "deep" approaches doesn't outperform "shallow" information extraction approaches (Yao et al. 2014)[1].

In many cases, you need constrained language, domain-specific syntax to get decent performance. However, Reddy (2017)[2] experimented with general-purpose dependency parser and concluded: "Our experiments on Free917, WebQuestions and GraphQuestions semantic parsing datasets conclude that general-purpose syntax is more useful for semantic parsing than induced task-specific syntax and syntax-agnostic representations."

Evaluation Edit

Old and small datasets:

Newer, large datasets:

  • Free917 (Cai and Yates, 2013)[4]
  • WebQuestions (Berant et al. 2013)[5]

References Edit

  1. Yao, X., Berant, J., & Van Durme, B. (2014). Freebase QA: Information Extraction or Semantic Parsing?. ACL 2014, 82.
  2. Reddy, Siva. 2017. "Syntax-Mediated Semantic Parsing."
  3. Zettlemoyer, L. S., & Collins, M. (2007, June). Online Learning of Relaxed CCG Grammars for Parsing to Logical Form. In EMNLP-CoNLL (pp. 678-687).
  4. Cai, Q., & Yates, A. (2013, August). Large-scale Semantic Parsing via Schema Matching and Lexicon Extension. In ACL (1) (pp. 423-433).
  5. Berant, J., Chou, A., Frostig, R., & Liang, P. (2013, October). Semantic Parsing on Freebase from Question-Answer Pairs. In EMNLP (pp. 1533-1544).