Erik G. Nilsson

Design patterns for user interface for mobile applications

The topic of this paper is a collection of user interface (UI) design patterns for mobile applications. In the paper we present the structure of the patterns collection - the patterns are suggested solutions to problems that are grouped into a set of problem areas that are further grouped into three main problem areas - a structure which is valuable both as an index to identifying patterns to use, and it gives a fairly comprehensive overview of issues when designing user interfaces for mobile applications. To show the breadth of the patterns collection we present six individual problems with connected design patterns in some detail - each coming from different problem areas. They represent important and relevant problems, and are on different levels of abstraction, thus showing how patterns may be used to present problems and solutions on different levels of detail. To show the relevance and usefulness of the patterns collection for usability professionals with a mixed background, we present some relevant findings from a validation of the patterns collection. In addition to verifying the relevance and usefulness of the patterns collection, it also shows both expected and surprising correlations between background and perceived relevance and usefulness. One important finding from the validation is an indication that the patterns collection is best suited for experienced UI developers wanting to start developing mobile UIs. Using a patterns collection for documenting design knowledge and experience has been a mixed experience, so we discuss pros and cons of this. Finally, we present related work and future research.

Model-based user interface adaptation

Most work on model-based cross-platform user interface development is based on an assumption that the user interfaces on the different platforms should be as similar as possible. Much work on mobile user interfaces claim the opposite A¢â‚¬â€œ that user interfaces on a mobile platform should have features not applicable on a stationary one and vise versa. Exploiting contextual information in user interfaces on mobile equipment is a prime example of this. This paper focus on this dichotomy between common development and exploiting platform specific features (or having specialized versions) on each platform. Few or none of the existing model-based languages and tools for user interface development are able to combine these two needs. These aspects are initially very difficult to combine, but in the paper we present an approach that makes this possible. First we briefly present our modelling approach, we pinpoint some of the general differences between mobile and stationary user interfaces, and we present an approach to building such self adapting systems where the adaptation is handled by generic middleware. Our approach builds on component frameworks and variability engineering to achieve adaptable systems, and property modelling, architectural reflection and context monitoring to support dynamic self-adaptation. With this as a background we investigate how the presented modelling approach may be extended and combined with the adaptive architecture to facilitate mode l-based user interface adaptation. Finally, we present some more general principles for how model-based approaches may be used when developing adaptive user interfaces.

Ad Hoc Networks and Mobile Devices in Emergency Response – A Perfect Match?

In this paper we use findings from three empirical studies to analyze how the use of wireless ad hoc networks as part of an ICT solution for emergency response imposes requirements to the user interface of these solutions. The analysis starts by arguing that explicit details about the network used (like availability, coverage and connected nodes) should be visualized for the user and may be used by applications to obtain useful information. It continues by discussing requirements to user interfaces for local leaders and field workers, identifying cross-platform support as an important need for the leaders and supporting different modalities as an important need for field workers. These and other requirements are used as input to an analysis of challenges when developing these user interfaces, concluding that handling flexibility is essential. Finally, we turn around and look at ad hoc networks from a user interface perspective. In particular, we present requirements to ad hoc networks used in ICT solutions for emergency response, focusing on size, speed and providing awareness of network status through the nodes in the network themselves.

Combining Compound Conceptual User Interface Components with Modelling Patterns - A Promising Direction for Model-Based Cross-Platform User Interface Development

In this paper we examine why model-based user interface development languages and tools only have had a limited dissemination outside the research communities, and argue that there will be an increasing need for cross-platform user interface development in the future. To meet these needs, user interface development languages and tools must use new approaches. We examine some alternatives, and conclude that an approach based on pattern-based abstract compound user interface components as building blocks is the most promising. We describe this approach in some detail, and give an example showing how three quite different instantiations of one modelling pattern may be mapped to different running user interfaces using a number of mapping rules to two different implementation platforms with significant differences. Then we discuss what is needed for modelling languages and tools following the described approach to be successful and give some concluding remarks.

Generic functionality in user interfaces for emergency response

In this paper we use findings from a number of empirical studies involving different emergency response actors to identify shared or overlapping needs for user interfaces functionality. By analyzing the findings from these studies, we have identified 11 categories of functionality supporting shared needs, including functionality for handling incident information, logging facilities, and functionality for managing human resources and equipment. After presenting our research method, we give an overview of the identified categories of shared functionality. We also describe one of the categories, namely resource management, in some more detail including giving examples of concrete user interface functionality. We have validated the conclusions of our findings through observations and interviews in a training exercise. The validation supported our prediction that the exercise would not reveal major additional categories of functionality, and it also supplemented the earlier findings regarding which actors that need which categories of functionality. We conclude by discussing pros and cons of using generic solutions supporting shared functionality across emergency response actors.

User profiles for adapting speech support in the opera web browser to disabled users

In this paper we describe results from our work on adapting speech support in the Opera web browser to disabled users, through using available gross categories of equipment feature (screen presentation and program control) to categorize user and usage characteristics in user profiles. Allocation of users to equipment is based on characteristics of user and equipment, rather than on diagnostic categories. We have combined a number of approaches to investigate how users with different kinds of disabilities may benefit from speech support in the Opera web browser, and how the speech support may be adapted to enhance their utility of this support. After an introduction, we present the method used, the user profiles, and how different types of voice support may be adapted to different (combinations of) profiles. The latter includes both general requirements and suggestions for presentation formats and commands (voice or keyboard) for the profiles. The main conclusions from the work are that the voice support in the version of the Opera web browser that was used in the study is most suitable for people who have reading and writing disorders and that further development of voice support should focus on better adaptation for persons with motor disabilities. Most blind and visually impaired people in Norway already have access to specialized support, and do not need the voice support in Opera.

Using a patterns-based modelling language and a model-based adaptation architecture to facilitate adaptive user interfaces

To design usable mobile applications, exploiting context changes is of vital importance. The rapid context changes in a mobile setting cause the need for flexible and adaptive user interfaces that are multitasking and possibly exploiting multiple modalities. Implementing adaptive user interfaces requires expensive application-specific solutions. Reuse of this type of solutions is difficult or impossible. To make it viable to implement adaptive user interfaces for a broader range of applications, there is both a need for new architecture and middleware, and ways of constructing applications. In this paper, we show how a combination of a patterns-based modelling language using compound user interface components and mapping rules as building blocks, and a generic adaptive architecture based on components with ports and utility functions for finding the optimal configuration in a given situation, facilitates implementation of applications with adaptive user interfaces. First we briefly present our modelling approach, and the adaptive architecture including the generic middleware exploiting architecture models at runtime. With this as a background we show how the presented modelling approach may be combined with the adaptive architecture to facilitate modelbased user interface adaptation. Finally, we compare our approach with other approaches for realizing adaptive user interfaces, and we give some conclusions and directions for future research.

A case-based assessment of the FLUIDE framework for specifying emergency response user interfaces

In this paper, we report the results from assessing the FLUIDE Framework for model-based specification of user interfaces supporting emergency responders. First, we outline the special challenges faced when developing such user interfaces, and the approach used in the FLUIDE Framework to meet these challenges. Then we introduce the framework, including its two specification languages. Thereafter, we present the case addressing the specification of user interfaces for three existing emergency response applications. Based on these specifications, we discuss how well we succeeded, concluding that we were able to describe the applications in a comprehensive and understandable way taking similarities and difference between the applications into account. The language constructs function as intended, having two languages has proven valuable, and the specifications scale quite well.

Using Application Domain Specific Runtime Systems and Lightweight User Interface Models

This paper presents an application domain specific run-time system using a lightweight user interface model to combine advanced user interfaces with flexible configuration mechanisms. The system is called the VUIM — VRP User Interface Module, and is a generic user interface run time system for vehicle routing applications. It balances between being a very general, model-based, non-application domain specific UI development tool and being a turnkey application (ready to use, but with very limited abilities to be adapted to specific needs). The VUIM includes a class library with a set of generic, but application domain-specific UI building blocks. These building blocks are on a higher level than simple UI controls, but more generic than a dialog box or a window in an application. The class library is augmented with a run time system (generic application), that “interprets” a UI model. The VUIM exemplifies a novel approach for CADUI.