Elisabeth G. Todd

Focussing on a standard pattern form: the development and evaluation of MUIP

A framework for the tool, MUIP (Management of User Interface Patterns), has been developed based on the relevant literature and a survey of existing pattern tools. The framework supports the following features: pattern authoring, manipulating forces, browsing patterns, searching patterns, versioning and customising patterns, relating patterns, manipulating collections and importing or exporting patterns. Patterns are described using the standard pattern form (PLML). An enhanced version of PLML, called PLML vl.2, has been developed so that pattern contents can be organised more effectively. Guided by this framework, a specification of a comprehensive pattern management system for manipulating dissimilar pattern collections was developed and a prototype implemented accordingly. A formal evaluation confirmed the usefulness of the prototype.

Introducing students to UI patterns

User interface (UI) patterns have the potential to assist students of Human Computer Interaction (HCI) to learn about the principles of design. They can be used to guide students developing a conceptual model of an interface. We investigated how HCI students used a method for developing UI-pattern models and ascertained their views on different pattern forms (illustrated and narrative-only). Data was collected by observation, questionnaires and completed exercises. The findings indicate that students found UI patterns an acceptable medium for presenting UI information, were positive about their experience building UI-pattern models, and that patterns aided communication.

Validating user interface pattern languages

When considering using a UI pattern language three types of validation need to be considered: The validity of the individual patterns, the internal validation of the pattern language and the external validation of the pattern language. A set of six tests that a developer can use to determine the internal validity of a pattern language has been identified.

UMM: a maturity model for UI-pattern languages

The set of tests developed to assess the internal validity of a user interface (UI) pattern languages [27, 28] form the basis of the proposed UI pattern language maturity model (UMM). UMM uses the UI pattern modelling technique developed for describing example UIs [26]. By building these models a UI pattern languages structure can be improved - a generative process. The UI pattern language maturity model is evaluated by using it to rate a selection of general UI pattern languages, plus two versions of a pattern language developed specifically for teaching purposes.

The role of functional prototyping in model validation

Summary form only given. Within the KADS framework, a model of domain expertise is built prior to system implementation. Before making important design decisions, it is necessary to check that this model correctly represents the domain and its problem solving aspects. In this paper, a framework which employs functional prototyping for validating a model of expertise is described. Before building the prototype, an object-oriented knowledge representation model is developed based on the model of expertise. This knowledge representation model provides the structure of the prototype. Since mistakes may be made in the translation from one model to another, the knowledge representation model has to be verified to ensure that it accurately depicts the information in the model of expertise. The functional prototype can then be constructed and verified. Finally, the domain experts can test out the functional prototype to see whether it accurately models the domain. This framework was applied to the summer-autumn management domain. Two important areas were prototyped: feed budgeting and production level analysis.

Using the KADS methodology: expert systems in farm management

Farmer decision making is an important area of farm management research. An expert systems development methodology, KADS, provides a method for analyzing the decision making processes of farmers. The KADS method provides a library of generic task models to guide the knowledge acquisition process. KADS facilitates the differentiation of domain knowledge from procedural knowledge. The authors report on the application of KADS to develop a task model of the decision making processes used by four expert seasonal supply dairy partners over the summer-autumn period.<<ETX>>