Title :
Comparison of connectionist and traditional methods applied to phrase classification and grammaticality determination
Author :
Wyard, P.J. ; Warren, N.P.
Author_Institution :
Speech & Language Technol. Sect. British Telecom Labs., Ipswich, UK
fDate :
12/1/1992 12:00:00 AM
Abstract :
Two sentence classification experiments are described. In the first, the neural net HODYNE (Higher Order Dynamic Topology Net) is shown to be better than a keyword method at classifying banking phrases. In the second, three techniques are compared at determining the grammaticality of strings of syntactic tags from the LOB (Lancaster Oslo Bergen) corpus: a method using the statistics of tag pairs. HODYNE, and a multilayer perception HODYNE and the statistical method perform significantly better than the MLP
Keywords :
bank data processing; computational linguistics; natural languages; neural nets; pattern recognition; HODYNE; Higher Order Dynamic Topology Net; Lancaster Oslo Bergen corpus; MLP; banking phrases; grammaticality determination; keyword method; multilayer perception; neural net; phrase classification; sentence classification; statistical method; syntactic tags; tag pairs;
Journal_Title :
Radar and Signal Processing, IEE Proceedings F