Premium
LUIS—A logic for task‐oriented user interface specification
Author(s) -
Stary Chris,
Pasztor Ana
Publication year - 1995
Publication title -
international journal of intelligent systems
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.291
H-Index - 87
eISSN - 1098-111X
pISSN - 0884-8173
DOI - 10.1002/int.4550100204
Subject(s) - computer science , human–computer interaction , user interface , variety (cybernetics) , domain (mathematical analysis) , domain knowledge , knowledge representation and reasoning , representation (politics) , task (project management) , axiom , modular design , programming language , artificial intelligence , politics , political science , law , economics , mathematical analysis , geometry , mathematics , management
The complexity of user interface design is still steadily increasing because of an increasing variety of innovative interaction styles (e.g., pen‐based interaction) and because of an increasing orientation towards tasks as they are actually accomplished in work environments. In order to handle the resulting variety of knowledge that has to be taken into account for design, specification techniques have to rely on representation schemes for end user tasks and their organization, problem domain data, and interaction styles. Declarative knowledge representation techniques provide the capabilities to handle heterogeneous categories of knowledge as well as their integration. This article describes not only a conceptual but also a formal framework for task‐oriented, declarative user interface specification. the conceptual framework is based on a semantic network. It is formalized comprehensively through LUIS (Logical User Interface Specification). LUIS is a novel many‐sorted temporal specification logic. Using LUIS all static and dynamic relations between (1) end user tasks and problem domain data, (2) end user tasks and interaction styles (media and modalities), and (3) problem domain data and interaction styles, according to different levels of refinement can be captured. In particular, LUIS supports to stepwise refinement of knowledge to elementary data structures and basic functions. Moreover, in LUIS organizational constraints concerning the accomplishment of end user tasks can be represented by postulating axioms applying the modalities “First,” “Sometime in the future,” “Next,” and “More recently than.” Since each knowledge category corresponds to a particular sort of the specification logic, the designer can be provided with modular knowledge sources. They can be extended or modified for further developments in a flexible way. © 1995 John Wiley & Sons, Inc.