Extending computational grammars by learning

• Unité de recherche: PCRD EU
• Numéro de projet: 97.0119
• Titre du projet anglais: Extending computational grammars by learning

Mots-clé (Anglais)

Machine learning; NLP; computational grammars

Autre Numéro de projet (Anglais)

EU project number: FMRXCT980237

Programme de recherche (Anglais)

EU-programme: 4. Frame Research Programme - 10.1 Stimulation of training and mobility

Description succincte (Anglais)

See abstract

Partenaires et organisations internationales (Anglais)

Coordinator: Rijksuniversiteit Groningen (NL)

Résumé des résultats (Abstract) (Anglais)

The practice of implementing grammars on the computer has promoted the development of a number of natural language processing applications. These grammatical systems have been invariably handcrafted, hence time and labour consuming and requiring linguistic expertise. An alternative to such purely knowledge-based approaches could emerge from exploring the potential of machine learning methods to elaborate and improve automatically the available state-of-the-art grammar models. This is the focus of the LCG network.
The LCG network applies and evaluates several currently interesting techniques for machine learning of natural language to a common problem - learning noun-phrase syntax. These techniques are as follows: Instance-based Learning; Neural Networks; Genetic Algorithms; Symbolic Finite State Methods; EBL; ILP; Maximum Entropy. Their individual success rate on the common research problem is evaluated with respect to the two component tasks of noun phrase analysis (i.e. the degree to which a correct linguistic structure is been assigned) and recognition (using the standard measures of recall and precision). Thus, the network contributes to the knowledge of whether and which machine learning techniques are well suited to the task of language learning.
The central web site for the project (http://www.uia.ac.be) provides a summary of the project and the results from the application of the research methods at the partner sites.

Références bases de données (Anglais)

Swiss Database: Euro-DB of the
State Secretariat for Education and Research
Hallwylstrasse 4
CH-3003 Berne, Switzerland
Tel. [phone]
Swiss Project-Number: 97.0119