|Data-Oriented Parsing (DOP) models of natural language, among others, embody the assumption that
human language perception and production work with representations of
past language experiences, rather than with abstract grammatical rules.
The DOP models operate by decomposing the given treebank representations into
fragments and recomposing those pieces to analyze new utterances.
||Lexical Functional Grammar (LFG) is a constraint-based theory of grammar. An LFG representation comprises constituent structure (c-structure) and functional structure (f-structure), along with links associating nodes in the c-structure with units of the f-structure.
The LFG-DOP model assumes a corpus annotated with LFG representations, decomposes these representations into LFG fragments and then recomposes them to provide LFG analyses for previously-unseen utterances.