Chapter 5
Data and methods
Identifying constructions bottom-up in the International Corpus of English
Article outline
- 5.1The International Corpus of English
- 5.2Data preparation and extraction
- 5.2.1Tagging and cleaning the corpora
- 5.2.2Data extraction
- 5.3Data annotation: Establishing constructions across levels of abstraction
- 5.3.1What counts as an argument?
- 5.3.1.1Noun phrase constituency
- 5.3.1.2Argument or Adjunct?
- 5.3.2Argument Structure Construction annotation
- 5.3.3Establishing the Lexically-Bound Constructions of make
: Adopting and adapting the CPA procedure
- 5.3.3.1The procedure in brief
- 5.3.3.2Valency patterns: Motivating non-canonical patterns as LBCs
- 5.3.3.3Motivating and defining the reduced set of semantic types
- 5.3.3.4Motivating the addition of semantic types in a data-driven fashion
- 5.3.3.5Motivating contextual roles in valency slots
- 5.3.3.6Motivating lexically specified items in valency slots
- 5.3.3.7Motivating subvalency features
- 5.3.3.8Diverging from the CPA methodology
- 5.4Conclusion
-
Notes