Improving and Facilitating the Placement of Interactive Elements on User Interfaces

Abstract

A user interface is the primary mean by which a user interacts with a computer. Interactive elements, placed on an interface, define the scope of interactions afforded to users. This thesis investigates placement issues central to the design of user interfaces. The primary goal is support the construction of user interfaces by improving or facilitating the placement of interactive elements on (1) graphical user interfaces (GUIs), and (2) post-WIMP user interfaces.

GUIs are the most-commonly used method for interacting with computers. They consists of interactive elements organised in a visual interface layout. Improving the construction of interface layouts positively impacts user performance and perception of the interface. However, objectively improving the placement of elements is non-trivial. The first part of my thesis addresses challenges towards improving placement on GUI layouts. To this end, I make two main contributions. In Sketchplore, I investigate design-time improvements by enabling interface designers to sketch and explore layouts using an interactive optimiser. In Familiarisation, I discuss a use-time approach to improve placement for individual users by applying principles of familiarity.

Post-WIMP interfaces go beyond the GUI paradigm, and open up new interaction possibilities. They support a larger set of interactive elements, such as sensors and actuators. Due to the added technical complexity, it can be challenging to place interactive elements onto such interfaces. The second part of this thesis focuses on facilitating the placement of interactivity onto post-WIMP interfaces, and makes two contributions towards addressing placement challenges. I investigate the placement of interactive electronic elements onto physical interfaces. I present PaperPulse as a tool for non-expert to place electronics onto paper interfaces, and extend the discussion to other physical interfaces such as wearables and smart home interfaces. In BinPut, I discuss the placement of standard input controls onto a diverse set of interfaces, and present a universal technique that can be applied to different types of input and devices.

The concepts and principles discussed in the thesis contribute towards addressing placement problems central to the construction of user interfaces. They can result in design interfaces that are performant, and that support a wide range of interactions. Quantitative and qualitative evaluations of the resulting tools and techniques provide evidence for the approaches presented in this dissertation.