Skip to Main content Skip to Navigation

Discourse Modeling with Abstract Categorial Grammars

Aleksandre Maskharashvili 1, 2
2 SEMAGRAMME - Semantic Analysis of Natural Language
Inria Nancy - Grand Est, LORIA - NLPKD - Department of Natural Language Processing & Knowledge Discovery
Abstract : This dissertation addresses the questions of discourse modeling within a grammatical framework called Abstract Categorial Grammars (ACGs). ACGs provide a unified framework for both syntax and semantics. We focus on the discourse formalisms that make use of a grammatical approach to capture the discourse structure regularities. In particular, we propose ACG encodings of two discourse formalisms: G-TAG and D-STAG. These ACG encodings shed light on the problem of clause-medial connectives that the G-TAG and D-STAG grammars leave out of account. Both G-TAG and D-STAG make use of an extra-grammatical processing to deal with discourse connectives that appear at clause-medial positions. In contrast, the ACG encodings of G-TAG and D-STAG offer a purely grammatical approach to clause-medial connectives. Each of these ACG encodings are second-order. Grammars of this class have reversibility properties that allow us to use the same polynomial algorithmes both for the discourse parsing and generation tasks.
Document type :
Complete list of metadata

Cited literature [46 references]  Display  Hide  Download
Contributor : Abes Star :  Contact
Submitted on : Friday, May 26, 2017 - 2:45:08 PM
Last modification on : Wednesday, November 3, 2021 - 7:08:53 AM
Long-term archiving on: : Monday, August 28, 2017 - 4:16:49 PM


Version validated by the jury (STAR)


  • HAL Id : tel-01412765, version 2


Aleksandre Maskharashvili. Discourse Modeling with Abstract Categorial Grammars. Computation and Language [cs.CL]. Université de Lorraine, 2016. English. ⟨NNT : 2016LORR0195⟩. ⟨tel-01412765v2⟩



Les métriques sont temporairement indisponibles