Research Into the Use of Speech Recognition Enhanced Microworlds in an Authorable Language Tutor
Abstract
An earlier ARI sponsored MILT project was designed to investigate the possibility of using natural language processing (NLP) software to identify semantic and syntactic errors and provide the basis for state of the art dialogue exercises. One of the thirteen exercise types developed was the microworld exercise. A microworld is a software environment in which students can issue commands that are executed by animation routines in a game like atmosphere. Once the first microworld exercise was completed and integrated into MILT, ARI funded the investigation of the use of discreet speech recognition technology in language learning using the microworld exercise as a basis. The goal of this current effort was to expand the capabilities of MILT and incorporate continuous speech recognition for Arabic, Spanish and English. The overall objective of this project was to develop a general purpose, authorable, microworld that utilizes continuous speech recognition. The central tasks were: (1) the design of an enhanced microworld exercise, (2) development of continuous speech recognition components for English, Arabic, and Spanish, (3) incorporation of speech recognition into the microworld exercise, and (4) expansion of the Arabic natural language processing (NLP) system.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 01, 1999
- Accession Number
- ADA362359
Entities
People
- A. Hamilton
- Beth Plott
- Eric Princen
- John Morgan
- Stephen Larocca