Project-Alpage:TEXT-ELABORATOR (2008

Team Alpage

Members

Overall Objectives

Scientific Foundations

From programming languages to linguistic grammars
Metagrammars
Symbolic parsing techniques
Probabilistic parsing approaches
Dynamic wide coverage lexical resources
Treebanks development and exploitation
Building and evaluating full-featured parsing systems
Standardization
Discourse structures
Coreference resolution

Application Domains

Panorama
Information extraction and knowledge acquisition
Processing answers to open-ended questions in surveys: vera
Generation of textual reports about statistical data: EASYTEXT
Processing e-mails: Kwaga

Software

Syntax
SxLfg
System DyALog
Tools and resources for Meta-Grammars
The Bonzai PCFG-LA parser
Alpage's linguistic workbench, including SxPipe
MElt
The syntactic lexicon Lefff and the Alexina framework
System EasyRef

New Results

Mildly Context-Sensitive formalisms
Membrane Grammars and Bracketed Contextual Grammars
Finite-State Multi-Tape Transducers
Automatic tools for developping, improving and correcting lexical resources
Merging syntactic lexical resources for improving the Lefff: continued
Comparing lexical resources for parsing: Lexicon-Grammar vs. Lefff
The Leffe and SpMg: a lexicon and a parser for Spanish
PerLex, a morphological lexicon for the Persian language
Lexconn: French Lexicon of Discourse Connectives
Designing efficient parsers using Meta-Grammars and DyALog
Probabilistic TIG-based dependency parsing
Optimized reduction of probabilized shared parse forests
Introducing beam search techniques in the Earley algorithm
Dependency parsing
Improving the lexical coverage of statistical parsers
Comparing various models for statistical parsing
Large scale corpus processing
State-of-the-art French tagging with MElt
Towards a better understanding of frequency effects in syntax
Discourse Synchronous TAGs: a formalism for discourse analysis
Processing of temporal information in French texts
Coreference Resolution
Sapiens: visualizing quotations in news wires

Contracts and Grants with Industry

TEXT-ELABORATOR (2008–2009)
Kwaga (ARITT contract, 2009)

Other Grants and Activities

National Initiatives
European Initiatives
International Initiatives
Exterior research visitors

Dissemination

Animation at INRIA and University Paris 7
Supervising
Committees
Participation to workshops, conferences, and invitations
Teaching

Bibliography

Major publications
Publications of the year
References in notes

Inria / Raweb 2009
Presentation of the Project Alpage

Section: Contracts and Grants with Industry

TEXT-ELABORATOR (2008–2009)

Participant : Laurence Danlos.

TEXT-ELABOATOR is a NLG (Natural Language Generation) project funded by TNS Sofres. It is leaded by the startup Watch System Assistance for whom L. Danlos works as a scientific consultant. The NLG system is operational within TNS since the fall of 2009.

Logo Inria