This thesis studies the creation of illusion in computer aided performance. Illusion is created here by using deceptions, and a design framework is presented which suggests several different deception strategies which may be useful. The framework has been developed in an iterative process in tandem with the development of 3 real world performances which were used to explore deception strategies. The first case study presents a system for augmenting juggling performance. The techniques that were developed to control this system demonstrate how deception may become useful even when the core of the performance is not deceptive in any way. This is followed by a magic performance called the Cup Game, which was designed to explicitly test the strategies of deception described in the framework. The final case study is an interactive art installation which presents the illusion of a pet rock that lives in a cage. This demonstrates the usefulness of suspension of disbelief in the creation of illusions. It also demonstrates interesting social effects that are used to strengthen this suspension of disbelief. The idea of creating the impression of a false situation is inspired particularly by previous HCI work on public interaction. This work demonstrated the usefulness of hiding interface use or computer outputs from some people in a situation. The creation of deliberately ambiguous computer interfaces, which allow for a wider variety of interpretations to be made by the user has also been described. The work here goes beyond these techniques to use technology to actively create false impressions. The techniques used in this process are guided by the work of magic performers, and by psychological studies of how magic performance works. As well as artistic performance, it is envisaged that this work may prove applicable to more traditional situations. In addition to the framework itself, the development of the case studies has created several useful algorithms which have wider applications. The case studies are also useful guides for those creating performance systems, or other systems where deceptive techniques may be useful.

Mobile mixed-reality experiences mix physical and digital spaces, enabling participants to simultaneously inhabit a shared environment online and on the streets. These experiences take the form of games, educational applications and new forms of performance and art, and engender new opportunities for interaction, collaboration and play. As mobile mixed-reality experiences move out of the laboratory and into more public settings they raise new challenges concerning how to support these experiences in the wild. This thesis argues that mobile mixed-reality experiences in which artists retain creative control over the content and operation of each experience, particularly those that are deployed as theatrical performances, require dedicated support for content authoring and reactive orchestration tools and paradigms in order to be successfully and robustly operated in public settings. These requirements are examined in detail, drawing on the experience of supporting four publicly toured mobile mixed-reality experiences; Can You See Me Now?, Uncle Roy All Around You, I Like Frank in Adelaide and Savannah, which have provided a platform to practically develop, refine and evaluate new solutions to answer these challenges in the face of presenting the experiences to many thousands of participants over a four year period. This thesis presents two significant supporting frameworks. The ColourMaps system enables designers to author location-based content by directly colouring over maps; providing a simple, familiar and yet highly flexible approach to matching location-triggers to complex physical spaces. It provides support for multiple and specialised content layers, and the ability to configure and manage other aspects of an experience, including filtering inaccurate position data and underpinning orchestration tools. Second, the Orchestration framework supports the day-to-day operation of public experiences; providing dedicated control-room tools for monitoring that reveal the content landscape and historical events, intervention and improvisation techniques for steering and shaping each participant's experience as it unfolds both physically and virtually, and processes to manage a constant flow of participants.

This research proposes a new architecture for Adaptive Educational Hypermedia Systems (AEHS). Architectures in the context of this thesis refer to the components of the system and their communications and interactions. The architecture addresses the limitations of AEHS regarding interoperability, reusability, openness, flexibility, and limited tools for collaborative and social learning. It presents an integrated adaptive and collaborative Web-based learning environment. The new e-learning environment is implemented as a set of independent Web services within a service-oriented architecture (SOA). Moreover, it uses a modern Learning Management System (LMS) as the delivery service and the user interface for this environment. This is a two-way solution, whereby adaptive learning is introduced via a widely adopted LMS, and the LMS itself is enriched with an external - yet integrated - adaptation layer. To test the relevance of the new architecture, practical experiments were undertaken. The interoperability, reusability and openness test revealed that the user could easily switch between various LMS to access the personalised lessons. In addition, the system was tested by students at the University of Nottingham as a revision guide to a Software Engineering module. This test showed that the system was robust; it automatically handled a large number of students and produced the desired adaptive content. However, regarding the use of the collaborative learning tools, the test showed low levels of such usage.

This thesis investigates two major classes of Evolutionary Algorithms, Genetic Algorithms (GAs) and Evolution Strategies (ESs), and their application to the Orthogonal Packing Problems (OPP). OPP are canonical models for NP-hard problems, the class of problems widely conceived to be unsolvable on a polynomial deterministic Turing machine, although they underlie many optimisation problems in the real world. With the increasing power of modern computers, GAs and ESs have been developed in the past decades to provide high quality solutions for a wide range of optimisation and learning problems. These algorithms are inspired by Darwinian nature selection mechanism that iteratively select better solutions in populations derived from recombining and mutating existing solutions. The algorithms have gained huge success in many areas, however, being stochastic processes, the algorithms' behaviour on different problems is still far from being fully understood. The work of this thesis provides insights to better understand both the algorithms and the problems. The thesis begins with an investigation of hyper-heuristics as a more general search paradigm based on standard EAs. Hyper-heuristics are shown to be able to overcome the difficulty of many standard approaches which only search in partial solution space. The thesis also looks into the fundamental theory of GAs, the schemata theorem and the building block hypothesis, by developing the Grouping Genetic Algorithms (GGA) for high dimensional problems and providing supportive yet qualified empirical evidences for the hypothesis. Realising the difficulties of genetic encoding over combinatorial search domains, the thesis proposes a phenotype representation together with Evolution Strategies that operates on such representation. ESs were previously applied mainly to continuous numerical optimisation, therefore being less understood when searching in combinatorial domains. The work in this thesis develops highly competent ES algorithms for OPP and opens the door for future research in this area.

This thesis is concerned with the uses of ubiquitous computing (ubicomp) technologies in everyday domestic settings. In contrast to goal-oriented technology adoption (e.g. in the workplace), the integration of pervasive technology in the home faces not only social but also physical and technical constraints. We propose a design framework for the introduction of ubicomp technology into today’s homes that, firstly, considers a holistic approach to integrating pervasive technology; secondly, takes into account social factors and domestic activity when defining the nature of the system’s interaction; and thirdly, allows the user to adapt the system’s interaction and collaboration. Most of the work to date on domestic ubicomp takes the customization of domestic spaces for granted, presuming that the integration of sensing technologies can be accomplished to any required degree and usually assuming that context-aware systems have to be proactive, limiting users to the role of consumers of the system’s actions rather than allowing them a more participative or cooperative role. We have applied our framework to design a domestic ubicomp system to support parents with childcare in the home. The “Context-Aware Room” and the “The Parent-Child Companion Tool” prototypes are built to take account of the interaction between the social and physical and the social and digital contexts in order to address issues of integration of sensing technology, socially respectful collaboration and system adaptation. Two studies explore the potential social acceptance of the PChCT. The panel study considers parents’ overall perceptions of whether these kinds of ubicomp tools might help with parental tasks. The usability study considers the usefulness and usability of the PChCT. The results of the study reflect a positive attitude to the PChCT. Parents liked the collaborative resources and facilities to tailor collaboration. Further work might be done to assess how the system’s adaptation might fit within a wider context of user needs. Nevertheless, we argue that the use of our framework can lead to more socially acceptable ubicomp experiences in the home.

This dissertation is about effects and type theory. Functional programming languages such as Haskell illustrate how to encapsulate side effects using monads. Haskell compilers provide a handful of primitive effectful functions. Programmers can construct larger computations using the monadic return and bind operations. These primitive effectful functions, however, have no associated definition. At best, their semantics are specified separately on paper. This can make it difficult to test, debug, verify, or even predict the behaviour of effectful computations. This dissertation provides pure, functional specifications in Haskell of several different effects. Using these specifications, programmers can test and debug effectful programs. This is particularly useful in tandem with automatic testing tools such as QuickCheck. The specifications in Haskell are not total. This makes them unsuitable for the formal verification of effectful functions. This dissertation overcomes this limitation, by presenting total functional specifications in Agda, a programming language with dependent types. There have been alternative approaches to incorporating effects in a dependently typed programming language. Most notably, recent work on Hoare Type Theory proposes to extend type theory with axioms that postulate the existence of primitive effectful functions. This dissertation shows how the functional specifications implement these axioms, unifying the two approaches. The results presented in this dissertation may be used to write and verify effectful programs in the framework of type theory.

This thesis describes a multi-agent approach to generating narrative based on the activities of participants in large-scale persistent virtual environments, such as massivelymultiplayer online role-playing games (MMORPGs). These environments provide diverse interactive experiences for large numbers of simultaneous participants. Involving such participants in an overarching narrative experience has presented challenges due to the difficulty of incorporating the individual actions of so many participants into a single coherent storyline. Various approaches have been adopted in an attempt to solve this problem, such as guiding players to follow pre-designed storylines, or giving them goals to achieve that advance the storyline, or by having developers (‘dungeon masters’) adapt the narrative to the real-time actions of players. However these solutions can be inflexible, and/or fail to take player interaction into account, or do so only at the collective level, for groups of players. This thesis describes a different approach, in which embodied witness-narrator agents observe participants’ actions in a persistent virtual environment and generate narrative from reports of those actions. The generated narrative may be published to external audiences, e.g., via community websites, Internet chatrooms, or SMS text messages, or fed back into the environment in real-time to embellish and enhance the ongoing experience with new narrative elements derived from participants’ own achievements. The design and implementation of this framework is described in detail, and compared to related work. Results of evaluating the framework, both technically, and through a live study, are presented and discussed.

This thesis is about Martin-Löf's intuitionistic theory of types (type theory). Type theory is at the same time a formal system for mathematical proof and a dependently typed programming language. Dependent types are types which depend on data and therefore to type check dependently typed programming we need to perform computation(normalisation) in types. Implementations of type theory (usually some kind of automatic theorem prover or interpreter) have at their heart a type checker. Implementations of type checkers for type theory have at their heart a normaliser. In this thesis I consider type checking as it might form the basis of an implementation of type theory in the functional language Haskell and then normalisation in the more rigorous setting of the dependently typed languages Epigram and Agda. I investigate a method of proving normalisation called Big-Step Normalisation (BSN). I apply BSN to a number of calculi of increasing sophistication and provide machine checked proofs of meta theoretic properties.

Virtual Environments offer the potential for users to explore social situations and experience different behaviour responses for a variety of simulated social interactions. One of the challenges for the VE developer is how to construct the VE to allow freedom of exploration and flexibility in interactive behaviour, without the risk of users deliberately or inadvertently missing important learning goals. The program has to be structured to guide the user in their learning and to take into account different levels of ability. This embedded ‘scaffolding’ within the VE software can aid the user’s learning in different contexts, such as individual, tutored or group learning situations. This thesis looks at the design and implementation of desktop VEs in a classroom for teaching social skills to people with Asperger’s Syndrome (AS). The first part of the thesis looks at work carried out as part of the AS Interactive project, a multidisciplinary research project using User Centred Design principles. VEs developed with the help of teachers and users were constantly refined in an iterative design process with evaluations and observations of the use of the VEs in the classroom to assess the effectiveness of elements used to scaffold the VEs. The last part of the thesis looks at work continued by the author after the end of the AS Interactive project after recommendations in that project for the VEs to fit the needs of the individual. Individualisation is researched with a number of demonstration and prototype VEs developed to help obtain information from autism experts and teachers on how best to individualise a learning VE for people with autism. The outcomes of this thesis include an exploration of the role of the programmer within a multi-disciplinary research group and the iterative development of VEs. A number of recommendations on how to scaffold VEs and make them usable in the classroom are then made. Finally recommendations are made on features and scenarios that could be useful in individualised learning VEs for people with autism and which require further evaluation in a classroom.

Measurement and monitoring are important for correct and efficient operation of a network, since these activities provide reliable information and accurate analysis for characterizing and troubleshooting a network’s performance. The focus of network measurement is to measure the volume and types of traffic on a particular network and to record the raw measurement results. The focus of network monitoring is to initiate measurement tasks, collect raw measurement results, and report aggregated outcomes. Network systems are continuously evolving: besides incremental change to accommodate new devices, more drastic changes occur to accommodate new applications, such as overlay-based content delivery networks. As a consequence, a network can experience significant increases in size and significant levels of long-range, coordinated, distributed activity; furthermore, heterogeneous network technologies, services and applications coexist and interact. Reliance upon traditional, point-to-point, ad hoc measurements to manage such networks is becoming increasingly tenuous. In particular, correlated, simultaneous 1-way measurements are needed, as is the ability to access measurement information stored throughout the network of interest. To address these new challenges, this dissertation proposes OverMon, a new paradigm for edge-to-edge network monitoring systems through the application of overlay techniques. Of particular interest, the problem of significant network overheads caused by normal overlay network techniques has been addressed by constructing overlay networks with topology awareness - the network topology information is derived from interior gateway protocol (IGP) traffic, i.e. OSPF traffic, thus eliminating all overlay maintenance network overhead. Through a prototype that uses overlays to initiate measurement tasks and to retrieve measurement results, systematic evaluation has been conducted to demonstrate the feasibility and functionality of OverMon. The measurement results show that OverMon achieves good performance in scalability, flexibility and extensibility, which are important in addressing the new challenges arising from network system evolution. This work, therefore, contributes an innovative approach of applying overly techniques to solve realistic network monitoring problems, and provides valuable first hand experience in building and evaluating such a distributed system.