Skip to Main content Skip to Navigation
New interface
Master thesis

Parsing Punctuation and Coordination Extragrammatically

Valmi Dufour-Lussier 1 
1 ORPAILLEUR - Knowledge representation, reasonning
INRIA Lorraine, LORIA - Laboratoire Lorrain de Recherche en Informatique et ses Applications
Abstract : Coordination is a syntactic construction that is extremely frequent in natural language and yet very difficult to analyse: it is highly ambiguous, as different types of constituents and non-constituents can be coordinated in different contexts, and it cannot be easily modelled using the same formal tools used to represent the "basic", coordination-less part of natural languages. As for punctuation, often neglected as an object of study by linguists because it is peculiar to written language, it can combine with or substitute conjunctions to play a coordinative role, or play an adjunctive role. Some researchers propose that coordination is not a grammatical phenomenon that is a matter of linguistic competence, but rather a performance issue that should be analysed directly among the syntactic structures. We suggest to use this idea to the benefit of natural language processing by defining an algorithm that deals with coordination and punctuation using graph transformations applied directly to the output of a parser based on a model of "basic" language. The syntactic structure of a sentence as proposed by our system takes the shape of a directed acyclic graph in which the constituent sharing phenomenon at the roots of coordination appears sharply. We detail an algorithm working within the framework of interaction grammars (but suggest ways to adapt it to other formalisms, namely tree-adjoining grammars, phrase structure grammars, and dependency syntax) which is able to parse many types of coordinative and adjunctive constructions.
Complete list of metadata
Contributor : Valmi Dufour-Lussier Connect in order to contact the contributor
Submitted on : Saturday, October 22, 2011 - 8:42:49 PM
Last modification on : Saturday, June 25, 2022 - 7:39:54 PM


  • HAL Id : inria-00634736, version 1



Valmi Dufour-Lussier. Parsing Punctuation and Coordination Extragrammatically. Artificial Intelligence [cs.AI]. 2010. ⟨inria-00634736⟩



Record views