In this presentation, Davina Jackson will:
• Give a general background about emerging policies from the Rudd Government of relevance to the Faculty of Architecture, Design and Planning and design research entities in Australia;
• Explain what’s happening with the emerging d_city global online research network, its main research visions and its potential points of relevance to the University’s researchers;
• Highlight the Metropolis Connecting Cities congress to be held in Sydney in October 2008 and opportunities for researchers to inform the audience of around 1000 mayors and senior government delegates from over 100 cities of more than a million people;
• Focus on the proposed d_city Video Lounge at the Metropolis Congress and potential to contribute videos of research relevant to data cities.
• Outline six global brands emerging from Sydney offering multi-disciplinary urban design leadership.
• How to involve creative output from the Faculty in one or more of these ventures.

Key words: data, city, dynamic digital design, architecture, research network, modelling, urban design, artificial intelligence.

SPEAKER BIOGRAPHY
Davina Jackson [M.Arch (history and theory) UNSW, PhD candidate RMIT] is an international writer, editor, occasional architectural historian and an entrepreneurial promoter of progressive architecture and design. A former political and crime reporter from New Zealand, she was the architecture editor for Vogue Living and Belle in the 1980s, editor of Architecture Australia in the 1990s, author of Australian Architecture Now (2000) and Next Wave: Emerging Talents in Australian Architecture (2007) as well as a guest critic and nominator for Phaidon’s World Atlases of Contemporary Architecture and 10X10_2 survey of 100 architects chosen by 10 international critics. She curated the Synthesis: Art + Architecture exhibition program in 1992 and 40UP: Australia’s Next Generation in Architecture (1999 Sydney, Melbourne; 2000 Hamburg, Berlin). From 2003-2005 she chaired a task force which successfully brokered a viable funding system for Australia to exhibit regularly at its often vacant national pavilion at Venice Architecture Biennales. From 2005-2007 she was Associate Professor of New South Global Design, advising three faculties at the University of New South Wales on emerging opportunities for cross-disciplinary urban design collaborations. Since 2005 she has been working with a steering committee of eminent digital architecture and design academics to establish a cross-disciplinary network of researchers in all fields relevant to the current revolution in data modelling future cities. She is now a Director of DCity Pty Ltd (organising an online presence to serve the d_city research community) and the Sydney International Smart Light Festival Ltd (organising a new biannual festival of low-energy urban lighting installations around the Sydney waterfront).

The goal of this research is to systematically measure the potential for sonification design to facilitate communication and interpretation of continuous quantitative data. Comprehensive investigation of the function of complex auditory displays in this study, attempts to bridge the gap between objective computational auditory events and subjective individual responses, and tie auditory display work together with perception-cognition-action relationships.

This paper presents case studies of the calculation of appraisals, the linguistic construal of emotions and attitudinal positions, in natural language design text. Using two data sets, one a standard set of movie review text and one a set of a natural language design text, we compare the performance of support vector machines for the classification of design documents by overall semantic orientation based on two different numerical representations of the design text. For the natural language design text data set, we additionally compare the performance of the support vector machine for the categorization of the text into three categories, Product, Process and People. We find that the sparse yet high dimensional representation of the design text allows the support vector machine to perform best. Further, we find modest benefit in encoding more knowledge about the semantic category and orientation into the representation beyond frequency counts on the occurrence of unigrams in the text.

Learning in design involves not only understanding design theories, or being exposed to the procedures, skills and processes necessary to design an object. It involves also understanding what counts as relevant and genuine within a particular design discipline, and how its existing group members recognise and realise legitimate design practices. This research examines how knowledge and identity are specialized within four design
disciplines: engineering, architecture, digital media and fashion design.
In this seminar, we present a Bersteinian approach to the development of the research's languages of description. A matrix used to guide the research was created in order to facilitate the mapping of the theory to empirical data and vice versa. Methodological implications are discussed through the lessons we have learned in the development of instruments (the interview questionnaire and the survey protocol), and the framework used to guide the analysis. We also present how Basil Bernstein's concepts of classification and framing and Karl Maton's concepts of 'legitimation codes', combined with the empirical results, might support instructional functions when embedded in an e-learning environment about experiencing the practice of design, in the informal learning setting of a museum.

The purpose of this research is to develop a conceptual framework and formal model that can help inform policy and guide investment in efforts to improve the quality of life for indigenous communities by exploring the fundamental capabilities required by indigenous communities to appropriate technologies in a manner they have reason to value. A meaningful measure of the relevant capabilities will be required to achieve the purpose.

This seminar will present unsupervised machine learning methods to automate design problem modeling and (re)-formulation in design optimization. The theoretical basis of developing these algorithms from a statistical pattern extraction perspective will be presented, i.e. how findings from cognitive science lie at the basis of the methodology development. Then, a knowledge-lean learning and inference mechanism based on Singular Value Decomposition (SVD) for design optimization problem (re)-formulation at the problem modeling stage will be presented. The distinguishing feature of the mechanism is that it requires very few training cases to extract and generalize knowledge for large classes of problems sharing similar characteristics. SVD is applied as a dimensionality reduction tool to extract semantic patterns from a syntactic formulation of the design problem. The mechanism is explained and evaluated on a model-based decomposition problem, a hydraulic cylinder design problem, and a medium-large scale Aircraft Concept Sizing problem. The results show that the method generalizes quickly and can be used to impute relations between variables, parameters, objective functions, and constraints when training data is provided in symbolic analytical form, and is likely to be extensible to forms when the representation is not in analytical functional form.

This seminar presents a computational framework for a model of interpretation as it relates to design thinking. Work has begun to implement a model that demonstrates the way that interpretation enables the use of experience to construct something meaningful from the data available. It looks at the balance between push from the data available and pull from experience. The computational framework uses a number of unsupervised category learners to represent memory within different levels of abstraction, with regions in each layer connected by auto-associative memory. Each layer within the model structures the data from the layer above it. A simple example with shapes is used for describing the framework.

The need for information anytime and anywhere has meant that computers have migrated from the familiar context of the desktop onto the human body in the form of wearable computers. Wearable visualizations or displays are the latest trends to emerge from the field of wearable computing, resulting in body-worn devices that represent data to the surrounding users. This research presents the development of two different wrist-worn displays that measure and convey skin conductivity, which in its broadest sense can be considered a measurement of physiological arousal in humans. Each distinctive display will vary in the level of comprehensibility and ease of understanding it affords to those who view it over a period of time. An in-depth evaluation study of usage patterns of each display will be conducted to assess the overall usability of wearable computing displays in every day contextual environments.

The aim of this research is to design and evaluate a Mixed Reality-supported collaborative system by considering the concept and characteristics of presence. This research introduces a Mixed Reality-supported collaborative system, called MR-Collab, to facilitate remote interior design by considering presence factors. It combines Mixed Reality (MR), more specifically Augmented Reality (AR), and state-of-the-art communication and collaboration technologies. The concept is partially based on Benford's Mixed Reality boundaries theory.
This system can facilitate collaboration between distributed designers and customers through Mixed Reality environment especially focus on measuring the degree of presence. The sense of "being there" and "we are together" could be achieved, which means MR-Collab allows users to experience the sense of being in the same "place" instead of being isolated while using conventional desktop computers . After the design and development of the system, the usability of the MR-Collab system will be evaluated and, the level of presence that generated when users are working with this system will be measured as well.

Nowadays, it is becoming more and more popular for teaching and learning to be supported in technology-supported settings, which could become new instructional methods. Tangible Augmented Reality (TAR) technology could create an innovative and interactive learning space by merging computer-generated elements of virtuality into a real learning space. It is envisaged that the combination of real and virtual media opens new perspectives for teaching and learning. However, different cognitive and social-learning processes are involved with different types of design learning activities. This Ph.D. proposal aims to evaluate Tangible Augmented Reality learning space for improving the pedagogical effectiveness of experiential and collaborative learning processes in urban design learning. There are three major components (framework, prototyping, and experimentation) in this proposal.
The theoretical framework part presented a theoretical foundation for designing and implementing TAR technologies to improve the pedagogical effectiveness of learning processes in different types of design learning activities. One TAR prototype is planned to be developed based on the justifications from the framework. The experimentation plans two experiments: one for validating TAR’s benefits in individual learning, the other for evaluate the effectiveness of TAR in social learning. The proposed time schedule is also included in this proposal.

This talk discusses a theory for the foundation of design for pedagogy based on the principles of pattern languages. Using this guiding theory, a variation of the Alexandrian pattern structure to embed pedagogy and teaching practices at the core of the design of e-learning. Based on pedagogical principles of constructivism and experiential learning this design for pedagogy pattern methodology addresses deficiencies in the use of patterns when applied to e-learning.

This talk presents a computational model for the construction of novel associations as a component of a larger project on analogy-making.
Association-construction is driven by a reinterpretation-based model of subjective similarity. Associations are constructed by transforming the way the agent perceives the objects being associated so that they become similar. A model is described where an agent develops associations by an iterative process of attempting to relate objects in its environment.

A computational model has been developed to study the role of different modes of social learning on the formation
team-mental-model(TMM) and team expertise. In this model, the TMM is formed over time as team members learn about each other. Members can learn about other team members and self from (a) their personal interactions with the task and other team members (b) observing other members perform tasks and interact with a third member. The different modes of learning are discretely represented, allowing controlled analysis of the role of each of the learning modes on formation of TMM.
The research framework, experiment design and simulation results used to validate the framework will be presented. Outcomes of this research have implications for the way teams are put together, how they are managed, and how they can be studied using computational simulations.

We are currently six months into a one year EPSRC funded feasibility study where we are investigating the usage of descriptive machine learning in graphic design applications. In particular, we are looking at using fitness function invention and closed-loop learning to attempt to model a user's aesthetic preferences in image filtering, and we are looking at inducing constraints from user examples in scene construction. In the talk, I will give a progress report and demonstrate some software we are developing for the project. I will use this to introduce a few notions from computational creativity that I have been developing recently.