文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > A COMPARISON OF NATURAL AND MACHINE LEARNING OF STRESS
侵权投诉

A COMPARISON OF NATURAL AND MACHINE LEARNING OF STRESS

In the language acquisition literature children are reported to make typical errors in assigning stress to words. These errors are taken to be unequivocal evidence that children learn rules for stress assignment as opposed to learning on a word-by-word bas

'LAZY LEARNING':A COMPARISON OF NATURAL AND MACHINE LEARNING OFSTRESS

Steven Gillis, Walter Daelemans, Gert Durieux

Abstract

In the language acquisition literature children are reported to make typical errors inassigning stress to words. These errors are taken to be unequivocal evidence that childrenlearn rules for stress assignment as opposed to learning on a word-by-word basis.

In this study an algorithm is used in a stress assignment task that incorporates exemplar-based learning as opposed to rule-based learning, viz. Instance-Based Learning (IBL).IBL belongs to a class of Lazy Learning algorithms which do not derive explicitabstractions such as rules, but instead basically memorize the learning material andgeneralize by analogical reasoning on the basis of examples stored in memory.

Experimental results show that IBL makes the same kinds of stress related errors inproduction that children make: (i) the amount of production errors is related to metricalmarkedness, and (ii) stress shifts and errors with respect to the segmental and syllabicstructure of words typically take the form of a regularization of stress patterns. We alsohighlight a type of error that is not extensively discussed in the language acquisitionliterature though it occurs quite frequently, viz. irregularizations. In an exemplar-basedmodel irregularizations can be elegantly explained by IBL's basic processingmechanisms.

上一页第2页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email