EPSRC logo

Details of Grant 

EPSRC Reference: GR/L81406/01
Title: ENHANCED LANGUAGE MODELLING THROUGH IMPROVED LEXICAL, GRAMMATICAL AND SYNTACTIC LABELLING
Principal Investigator: Huckvale, Professor MA
Other Investigators:
Researcher Co-Investigators:
Project Partners:
QinetiQ
Department: Phonetics and Linguistics
Organisation: UCL
Scheme: Standard Research (Pre-FEC)
Starts: 01 April 1998 Ends: 31 March 2001 Value (£): 172,895
EPSRC Research Topic Classifications:
Human Communication in ICT
EPSRC Industrial Sector Classifications:
No relevance to Underpinning Sectors
Related Grants:
Panel History:  
Summary on Grant Application Form
The overall aim of the research is the creation of measurable enhancements to the performance of a trigram language model for speech recognition through the improved use of linguistic information at lexical, grammatical and syntactic levels. Our proposal has three main novel elements:1. the introduction of a lexical analysis component that reduces data sparseness through analyses of the morphological structure and the collocational preference of words,2. the introduction of a rich word-class tagging system enhanced for the particular problems of statistical language modelling,3. the introduction of phrase level parsing based on the Survey parser constructed from the current EPSRC project. We aim to evaluate our designs on standard corpora using the standard measures of perplexity. We also propose to evaluate our results on N-best recognition results from a large-vocabulary speech recognition system.
Key Findings
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
Potential use in non-academic contexts
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
Impacts
Description This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
Summary
Date Materialised
Sectors submitted by the Researcher
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
Project URL:  
Further Information:  
Organisation Website: