Task-oriented parsing - a diagnostic method to be used adaptive systems

1 January 1988

proceedings article
Published by Association for Computing Machinery (ACM)

p. 241-247
https://doi.org/10.1145/57167.57208

Abstract

In order to be able to show context-dependent responses to the user's actual needs, adaptive systems have to be provided with models of possible task contexts. Existing methods for the representation of tasks in HCI are insufficient for this purpose as they do not support task-oriented parsing (i.e. analysing the input stream in terms of higher level task units). This paper presents a Prolog implementation of a task-oriented parser (+ generator) based on a grammar notation called LEXITAS. As an application, an online coach for a UNIX-like file management system is described. Further applications, such as automated macro detection from given interaction protocols, are discussed.

Keywords

POSSIBLE TASK CONTEXT
INTERFACE DESIGN TOOLS & TECH- NIQUES
TASK-ORIENTED PARSING
DIAGNOSTIC METHOD
ACTUAL NEED
ADAPTIVE SYSTEM
AUTOMATED MACRO DETECTION
PROLOG IMPLEMENTATION
ADAPTIVE SYSTEMS.
TASK-ORIENTED PARSER
HIGHER LEVEL TASK UNIT
INTERACTION LANGUAGES AND NOTATIONS
UNIX-LIKE FILE MANAGEMENT SYSTEM
CONTEXT-DEPENDENT RESPONSE
TASK & INTERACTION ANALYSIS
MANAGEMENT SYSTEM
CONTEXT DEPENDENT
INTERFACE DESIGN

This publication has 0 references indexed in Scilit: