We attack the problem of user fatigue by using an interactive genetic algorithm to evolve user interfaces in the XUL interface definition language. The interactive genetic algorithm combines a set of computable user interface design metrics with subjective user input to guide the evolution of interfaces. Our goal is to provide user interface designers with a tool that can be used to explore innovation and creativity in the design space of user interfaces and make it easier for end-users to further customize their user interface without programming knowledge. User interface specifications are encoded as individuals in an interactive genetic algorithm’s population and their fitness is computed from a weighted combination of user interface design guidelines and user input. This paper shows that we can reduce human fatigue in interactive genetic algorithms (the number of choices needing to be made by the designer), by 1) only asking the user to pick two user interfaces from among ten shown on the display and 2) by asking the user to make the choice once every t generations.

Graphical user interface design is a time consuming, expensive, and complex software design process. User interface design is both art and science in that we use both objective and subjective design metrics to evaluate interfaces. An automated process that relies on both subjective and objective metrics to guide the evolution of effective, personalized user interfaces could significantly change current GUI development and maintenance practice. This paper uses an interactive genetic algorithm to evolve XUL user interface layouts by combining objective and subjective metrics. The genetic algorithm encodes expert knowledge from prominent usability guidelines as objective heuristics. Further, the graphical user interface developer (or user!) biases and guides the evolution of the interfaces by subjectively evaluating and selecting the “best ” and “worst ” interfaces from a small set of displayed interface prototypes. We explore how the selection of individuals from the population to be displayed to the user for subjective evaluation affects the convergence of the genetic algorithm and show that our methodology can produce effective interfaces that reflect subjective user-preferred aesthetics.

We introduce the idea of partial behaviours in user interfaces. A partial behaviour can beguile users, and may be a cause of interaction problems — however, it is possible to identify and remove them early in the design process, making them a useful concept for interaction design. A characterisation of partial behaviours is presented, in terms of a matrix algebra model of interactive systems. We use the model to show some real interfaces have undesirable and apparently unnecessary partial behaviours, and we discuss how choices made when modelling affect our notion of partiality. We also briefly describe a design tool that provides automatic support for partial behaviour analysis.

Computer algebra systems can do impressive mathematics that can help enormously in certain areas of formal HCI. This paper shows the ability to generate formal specifications, explore and generate theorems relevant to HCI needs, and how to do this automatically and reliably from existing, straight-forwardly programmed runnable systems. Conventional iterative design can modify these implementations, and we can then automatically redo formal analyses as the design is updated.

The Cognitive Dimensions framework has inspired research both more and less varied than expected. In this paper we revisit the original aims and briefly describe some subsequent research, to consider whether the original aims were too austere in rejecting knowledge-based dimensions; whether the dimensions can be shown to have real-world relevance; and whether their definitions can be improved, either piecemeal or by refactoring the entire set. We mention some issues that remain unexplored, and conclude by describing two different ventures into defining clear procedures for real-life application, operating in very different milieux but both accepting that the framework should be developed from its original formulation. 2 1.

Abstract. We introduce a formal model of inconsistency-related mode confusion. This forms the basis of a heuristic methodology for critiquing user interfaces, using a matrix algebra approach to interface specification [11]. We also present a novel algorithm for automatically identifying modes in state-based interface designs, allowing a significant level of automated tool support for our methodology. The present paper extends and generalises our previous work on improving state-based interface designs [4].

We introduce the idea of partial behaviours in user interfaces. A partial behaviour can beguile users, and may be a cause of interaction problems — however, it is possible to identify and remove them early in the design process, making them a useful concept for interaction design. A characterisation of partial behaviours is presented, in terms of a matrix algebra model of interactive systems. We use the model to show some real interfaces have undesirable and apparently unnecessary partial behaviours, and we discuss how choices made when modelling affect our notion of partiality. We also briefly describe a design tool that provides automatic support for partial behaviour analysis.

Abstract. Usability evaluation methods are a battery of techniques for assessing the usability of interactive systems or of proposed interactive systems. This paper describes a new evaluation method, particularly appropriate for evaluating safety critical and high quality user interfaces. The method can also be used for informing HCI research. The method is applied when a specification is available of an interactive system, or when a system (or prototype) is working. 1

Calculators have traditional user interfaces that no longer fit with acceptable practice in user interface design. This paper proposes a method for evaluating calculator user interfaces and demonstrates its value on a radically different style of calculator. The method produces a strong correlation between prior competence of the user and the time taken to achieve tasks with a given calculator despite differences in interfaces. Using this method, new calculator designs could be effectively evaluated and hence lead to long term improvements in their design. 1 The problem with calculators In this paper, we consider basic four function calculators, that is, ones with the four arithmetic functions together with one or two extra facilities such as a memory, percent and square root. The user interfaces for such calculators has remained reasonably static for the last thirty years. This is exemplified by the two Casio calculators 101-F and HS-8V showing in Figure 1. The former was

Abstract. Papers in HCI play different roles, whether to inspire, solve industrial problems or further the science of HCI. There is a potential conflict between the different views, and a danger that different forms of validity are assumed by author and reader — deliberately or accidentally. This paper reviews some of the issues in this complex area and makes practical recommendations. In particular, the paper introduces the term “cross-validity ” to help make explicit the issues, problems and means to tackle them. 1