Chancellery Research - Research Publications

Permanent URI for this collection

Search Results

Now showing 1 - 10 of 585
  • Item
    No Preview Available
    On the impact of initialisation strategies on Maximum Flow algorithm performance
    Alipour, H ; Munoz, MA ; Smith-Miles, K (PERGAMON-ELSEVIER SCIENCE LTD, 2024-03)
  • Item
    No Preview Available
  • Item
    No Preview Available
    Directive Explanations for Actionable Explainability in Machine Learning Applications
    Singh, R ; Miller, T ; Lyons, H ; Sonenberg, L ; Velloso, E ; Vetere, F ; Howe, P ; Dourish, P (ASSOC COMPUTING MACHINERY, 2023-12)
    In this article, we show that explanations of decisions made by machine learning systems can be improved by not only explaining why a decision was made but also explaining how an individual could obtain their desired outcome. We formally define the concept of directive explanations (those that offer specific actions an individual could take to achieve their desired outcome), introduce two forms of directive explanations (directive-specific and directive-generic), and describe how these can be generated computationally. We investigate people’s preference for and perception toward directive explanations through two online studies, one quantitative and the other qualitative, each covering two domains (the credit scoring domain and the employee satisfaction domain). We find a significant preference for both forms of directive explanations compared to non-directive counterfactual explanations. However, we also find that preferences are affected by many aspects, including individual preferences and social factors. We conclude that deciding what type of explanation to provide requires information about the recipients and other contextual information. This reinforces the need for a human-centered and context-specific approach to explainable AI.
  • Item
    No Preview Available
    The Impact of Judgment Variability on the Consistency of Offline Effectiveness Measures
    Rashidi, L ; Zobel, J ; Moffat, A (ASSOC COMPUTING MACHINERY, 2024-01)
    Measurement of the effectiveness of search engines is often based on use of relevance judgments. It is well known that judgments can be inconsistent between judges, leading to discrepancies that potentially affect not only scores but also system relativities and confidence in the experimental outcomes. We take the perspective that the relevance judgments are an amalgam of perfect relevance assessments plus errors; making use of a model of systematic errors in binary relevance judgments that can be tuned to reflect the kind of judge that is being used, we explore the behavior of measures of effectiveness as error is introduced. Using a novel methodology in which we examine the distribution of “true” effectiveness measurements that could be underlying measurements based on sets of judgments that include error, we find that even moderate amounts of error can lead to conclusions such as orderings of systems that statistical tests report as significant but are nonetheless incorrect. Further, in these results the widely used recall-based measures AP and NDCG are notably more fragile in the presence of judgment error than is the utility-based measure RBP, but all the measures failed under even moderate error rates. We conclude that knowledge of likely error rates in judgments is critical to interpretation of experimental outcomes.
  • Item
    No Preview Available
    Generating Dynamic Kernels via Transformers for Lane Detection
    Chen, Z ; Liu, Y ; Gong, M ; Du, B ; Qian, G ; Smith-Miles, K (IEEE, 2023-01-01)
  • Item
    No Preview Available
    Near-Wall Flow Statistics in High-Reτ Drag-Reduced Turbulent Boundary Layers
    Deshpande, R ; Zampiron, A ; Chandran, D ; Smits, AJ ; Marusic, I (SPRINGER, 2023-01-01)
  • Item
    Thumbnail Image
    plyranges: a grammar of genomic data transformation
    Lee, S ; Cook, D ; Lawrence, M (BMC, 2019-01-04)
    Bioconductor is a widely used R-based platform for genomics, but its host of complex genomic data structures places a cognitive burden on the user. For most tasks, the GRanges object would suffice, but there are gaps in the API that prevent its general use. By recognizing that the GRanges class follows "tidy" data principles, we create a grammar of genomic data transformation, defining verbs for performing actions on and between genomic interval data and providing a way of performing common data analysis tasks through a coherent interface to existing Bioconductor infrastructure. We implement this grammar as a Bioconductor/R package called plyranges.
  • Item
    Thumbnail Image
    A preliminary investigation into the influence of archaeological material on the yellowing of polyethylene storage bags
    Thompson, K ; Nel, P (Routledge, 2021)
    Concerns around the degradation of plastics have been part of conservation discourse for decades. The spotlight is usually on art and objects, and conservation and display materials, however it could be argued that a significant volume of the plastics in museums is associated with storage bags. This study asked whether the condition of plastic storage bags might be influenced by what is stored inside them. If specific materials can be identified as more likely to affect plastic degradation, museums may have a lead-indicator for efficiently monitoring storage risks. This case study developed a methodology for applying multivariate analysis to collected data to answer this question. A subset of polyethylene self-seal bags used to pack archaeological material from the ‘Casselden Place’ assemblage at Museums Victoria was evaluated. Objective data were combined with subjective assessment of bag degradation features gathered during a collection survey and interrogated using multivariate statistical analysis. Results indicate (1) different levels of yellowing are associated with particular plastic bag stocks and (2) ceramic, slate and tile finds are more likely than other materials to be contained within yellower bags. The research points to future enquiry and demonstrates this methodology shows promise for extension to other large cultural datasets.
  • Item
    Thumbnail Image
    Pressure drag reduction via imposition of spanwise wall oscillations on a rough wall
    Deshpande, R ; Kidanemariam, AG ; Marusic, I (Cambridge University Press, 2024-01-11)
    The present study tests the efficacy of the well-known viscous drag reduction strategy of imposing spanwise wall oscillations to reduce pressure drag contributions in transitional and fully rough turbulent wall flow. This is achieved by conducting a series of direct numerical simulations of a turbulent flow over two-dimensional (spanwise-aligned) semi-cylindrical rods, placed periodically along the streamwise direction with varying streamwise spacing. Surface oscillations, imposed at fixed viscous-scaled actuation parameters optimum for smooth wall drag reduction, are found to yield substantial drag reduction ( $\gtrsim$ 25 %) for all the rough wall cases, maintained at matched roughness Reynolds numbers. While the total drag reduction is due to a drop in both viscous and pressure drag in the case of transitionally rough flow (i.e. with large inter-rod spacing), it is associated solely with pressure drag reduction for the fully rough cases (i.e. with small inter-rod spacing), with the latter being reported for the first time. The study finds that pressure drag reduction in all cases is caused by the attenuation of the vortex shedding activity in the roughness wake, in response to wall oscillation frequencies that are of the same order as the vortex shedding frequencies. Contrary to speculations in the literature, this study confirms that the mechanism behind pressure drag reduction, achieved via imposition of spanwise oscillations, is independent of the viscous drag reduction. This mechanism is responsible for weakening of the Reynolds stresses and increase in base pressure in the roughness wake, explaining the pressure drag reduction observed by past studies, across varying roughness heights and geometries.
  • Item
    No Preview Available
    Indigenous peoples' health after Australia's No vote
    Chamberlain, C ; Anderson, I ; Fredericks, B ; Calma, T ; Eades, S (BMJ Publishing Group, 2024-01-11)