Keynote Speakers

Professor Stephen Kisely

Health LinQ - Using routine health data to link to better patient care

Australia is one of few countries that have comprehensive, high-quality, population data on many aspects of health and health care. Administrative health data have several advantages over community surveys, or data derived from individual clinical settings. They provide accessible and timely longitudinal data for an entire jurisdiction without the intrusion and cost of additional data collection, and so can be useful for both research and chronic disease surveillance. The Australian Government has provided $30 million to establish the Population Health Research Network (PHRN), with representation from all States and Territories to facilitate population health research through data linkage. The Queensland Centre for Health Data Services (QCHDS) is the Queensland node and trades under Health LinQ. The centre involves 4 Queensland universities, Queensland Health and the Australian e-Health Research Centre. Functions of the QCHDS include: 1) Facilitating access to linked datasets involving Queensland Health and national data; 2) Developing methodologies for data linkage and analysis; 3) Capacity building around data linkage; 4) Participation in the national coordination of data linkage and research. This talk describes the procedures for, and applications of, health data linkage. Researchers can either define cohorts for study within the administrative data or link them to their own data. Linkages are by probabilistic and deterministic matching using iterations of Link King and Febrl software packages, with robust protocols to preserve patient confidentiality. Access to the following data has been achieved: hospital morbidity, mortality, peri-natal, mental health data, as well as vital statistics. Privacy of linked data is protected by using a broken chain of information.

Data custodians provide demographic information without any health service data for linkage. A unique key is attached to the health service data before being handed through Health LinQ to the researcher. The researcher now has access to linked data without receiving any of the identifying demographic information. Current projects include preventable deaths from physical illness in psychiatric patients and an evaluation of whether the increased tax on alcopops has reduced alcohol-related health service use such as admissions to hospital or visits to emergency departments.

Presenter: Steve Kisely is a physician whose research and clinical interests are in epidemiology/pharmaco-epidemiology, chronic disease surveillance, health services research (HSR), and physical & psychiatric co-morbidity. Steve was the winner of Special Judges Award in the category of Best Use of IT in Clinical Care in Great Britain as part of the 1998 National Health Care IT Effectiveness Awards. He is currently Director of the Queensland Centre for Health Data Services.

Professor Andy Cockburn

Revisiting the Human as an Information Processor

In their seminal 1983 book ´The Psychology of Human–Computer Interaction´, Card, Newell and Moran (ACM Fellows and Turing Award winners) introduced ‘the model human information-processor’. This model equipped interface designers with strong theoretical tools to predict human interface performance without the demands of implementation and evaluation.

In the quarter century since then, human–computer interaction research has been extremely successful in transitioning research ideas to commercial deployment, yet interface design remains something of an art that is dependent on time–consuming iterations of design, implement, and evaluate. Theoretical models of human performance are rarely used despite their potential.

In this presentation I will describe several of our recent projects seeking to improve the efficiency of everyday activities in computer use, including scrolling, text messaging, window switching, and navigating through menu and file structures. The overriding theme, however, is on using theoretical human performance models to inform design, explain and predict performance, and to generalise results obtained. Ultimately, the objective is to give all computer science graduates the equivalent of a ´Big O´ complexity theory for user interfaces that allows them to design with assurance.

Presenter: Andy Cockburn is a Professor of Computer Science and Software Engineering at the University of Canterbury, Christchurch, New Zealand. He is interested in most areas of Human–Computer Interaction, but particularly in improving every day activities of computer use such as navigating within and between documents, and in understanding fundamental human factors influencing user performance.

Professor Alon Halevy

Structured Data on the Web

Though the World Wide Web is mostly known for its vast collection of textual data, the Web also offers access to millions of structured databases in almost every domain imaginable. Leveraging these collections for better search raises some significant challenges. In addition, the Web enables novel opportunities for collaborative data creation, management and analysis. In this talk I will describe some of the projects at Google that address these challenges.

In one project we crawled the content of millions of databases behind forms and now serve content from these databases to over 1000 queries per second. In the second, I will describe what can be done with a collection of 150 million high-quality data tables, 5 orders of magnitude greater than any previous collection ever managed. Finally, I will describe novel tools for collaborative data creation and discussion.

Presenter: Alon Halevy heads the Structured Data Management Research group at Google. Prior to that, he was a professor of Computer Science at the University of Washington in Seattle. In 1999, Dr. Halevy co–founded Nimble Technology, one of the first companies in the Enterprise Information Integration space, and in 2004, Dr. Halevy founded Transformic Inc., a company that created search engines for the deep web, and was acquired by Google. Dr Halevy is a Fellow of the Association for Computing Machinery, received the the Presidential Early Career Award for Scientists and Engineers (PECASE) in 2000, and was a Sloan Fellow (1999–2000). He published over 150 technical papers. He received his Ph.D in Computer Science from Stanford University in 1993.