Computing and Information Systems - Research Publications
Permanent URI for this collection
Search Results
1 - 10 of 12
-
ItemUncovering Students’ Learning Pathways: A Process Mining Perspective( 2023)This paper presents an approach to discovering students’ pathways when accessing a Learning Management System (LMS). These pathways reflect students’ compliance with the subject design and/or alternate ways of learning. Discovering such routines can enable the early detection of students at risk of not achieving the intended learning outcomes, as well as informing academics about students’ understanding of their progression in the subject. While LMSs report on aggregate data, they do not report on the order in which students follow the subject design. This information can reveal undesirable situations, such as students responding to the quizzes before completing the prerequisite activities (e.g., watching videos or completing the readings).
-
ItemLearning When to Treat Business Processes: Prescriptive Process Monitoring with Causal Inference and Reinforcement Learning( 2023-03-06)Increasing the success rate of a process, i.e. the percentage of cases that end in a positive outcome, is a recurrent process improvement goal. At runtime, there are often certain actions (a.k.a. treatments) that workers may execute to lift the probability that a case ends in a positive outcome. For example, in a loan origination process, a possible treatment is to issue multiple loan offers to increase the probability that the customer takes a loan. Each treatment has a cost. Thus, when defining policies for prescribing treatments to cases, managers need to consider the net gain of the treatments. Also, the effect of a treatment varies over time: treating a case earlier may be more effective than later in a case. This paper presents a prescriptive monitoring method that automates this decision-making task. The method combines causal inference and reinforcement learning to learn treatment policies that maximize the net gain. The method leverages a conformal prediction technique to speed up the convergence of the reinforcement learning mechanism by separating cases that are likely to end up in a positive or negative outcome, from uncertain cases. An evaluation on two real-life datasets shows that the proposed method outperforms a state-of-the-art baseline.
-
ItemTrust and Medical AI: The challenges we face and the expertise needed to overcome them( 2020-08-18)Artificial intelligence (AI) is increasingly of tremendous interest in the medical field. However, failures of medical AI could have serious consequences for both clinical outcomes and the patient experience. These consequences could erode public trust in AI, which could in turn undermine trust in our healthcare institutions. This article makes two contributions. First, it describes the major conceptual, technical, and humanistic challenges in medical AI. Second, it proposes a solution that hinges on the education and accreditation of new expert groups who specialize in the development, verification, and operation of medical AI technologies. These groups will be required to maintain trust in our healthcare institutions.
-
ItemThe Three Ghosts of Medical AI: Can the Black-Box Present Deliver?( 2020-12-10)Our title alludes to the three Christmas ghosts encountered by Ebenezer Scrooge in \textit{A Christmas Carol}, who guide Ebenezer through the past, present, and future of Christmas holiday events. Similarly, our article will take readers through a journey of the past, present, and future of medical AI. In doing so, we focus on the crux of modern machine learning: the reliance on powerful but intrinsically opaque models. When applied to the healthcare domain, these models fail to meet the needs for transparency that their clinician and patient end-users require. We review the implications of this failure, and argue that opaque models (1) lack quality assurance, (2) fail to elicit trust, and (3) restrict physician-patient dialogue. We then discuss how upholding transparency in all aspects of model design and model validation can help ensure the reliability of medical AI.
-
ItemComputer Science Communities: Who is Speaking, and Who is Listening to the Women? Using an Ethics of Care to Promote Diverse Voices( 2021-01-18)Those working on policy, digital ethics and governance often refer to issues in `computer science', that includes, but is not limited to, common subfields of Artificial Intelligence (AI), Computer Science (CS) Computer Security (InfoSec), Computer Vision (CV), Human Computer Interaction (HCI), Information Systems, (IS), Machine Learning (ML), Natural Language Processing (NLP) and Systems Architecture. Within this framework, this paper is a preliminary exploration of two hypotheses, namely 1) Each community has differing inclusion of minoritised groups (using women as our test case); and 2) Even where women exist in a community, they are not published representatively. Using data from 20,000 research records, totalling 503,318 names, preliminary data supported our hypothesis. We argue that ACM has an ethical duty of care to its community to increase these ratios, and to hold individual computing communities to account in order to do so, by providing incentives and a regular reporting system, in order to uphold its own Code.
-
ItemNo Preview AvailableThe acceptability and uptake of smartphone tracking for COVID-19 in Australia(Center for Open Science, 2020)
In response to the COVID-19 pandemic, many Governments are instituting mobile tracking technologies to perform rapid contact tracing. However, these technologies are only effective if the public is willing to use them, implying that their perceived public health benefits must outweigh personal concerns over privacy and security. The Australian federal government recently launched the `COVIDSafe' app, designed to anonymously register nearby contacts. If a contact later identifies as infected with COVID-19, health department officials can rapidly followup with their registered contacts to stop the virus' spread. The current study assessed attitudes towards three tracking technologies (telecommunication network tracking, a government app, and Apple and Google's Bluetooth exposure notification system) in two representative samples of the Australian public prior to the launch of COVIDSafe. We compared these attitudes to usage of the COVIDSafe app after its launch in a further two representative samples of the Australian public. Using Bayesian methods, we find widespread acceptance for all tracking technologies, however, observe a large intention-behaviour gap between people’s stated attitudes and actual uptake of the COVIDSafe app. We consider the policy implications of these results for Australia and the world at large.
-
ItemEnhancing Utility in the Watchdog Privacy Mechanism( 2021-10-10)This paper is concerned with enhancing data utility in the privacy watchdog method for attaining information-theoretic privacy. For a specific privacy constraint, the watchdog method filters out the high-risk data symbols through applying a uniform data regulation scheme, e.g., merging all high-risk symbols together. While this method entirely trades the symbols resolution off for privacy, we show that the data utility can be greatly improved by partitioning the high-risk symbols set and individually privatizing each subset. We further propose an agglomerative merging algorithm that finds a suitable partition of high-risk symbols: it starts with a singleton high-risk symbol, which is iteratively fused with others until the resulting subsets are private.~Numerical simulations demonstrate the efficacy of this algorithm in privately achieving higher utilities in the watchdog scheme.
-
ItemProcess Mining-Driven Analysis of the COVID19 Impact on the Vaccinations of Victorian Patients( 2021-12-08)Process mining is a discipline sitting between data mining and process science, whose goal is to provide theoretical methods and software tools to analyse process execution data, known as event logs. Although process mining was originally conceived to facilitate business process management activities, research studies have shown the benefit of leveraging process mining tools in different contexts, including healthcare. However, applying process mining tools to analyse healthcare process execution data is not straightforward. In this paper, we report the analysis of an event log recording more than 30 million events capturing the general practice healthcare processes of more than one million patients in Victoria–Australia–over five years. Our analysis allowed us to understand benefits and limitations of the state-of-the-art process mining techniques when dealing with highly variable processes and large data-sets. While we provide solutions to the identified limitations, the overarching goal of this study was to detect differences between the patients‘ health services utilization pattern observed in 2020– during the COVID-19 pandemic and mandatory lock-downs –and the one observed in the prior four years, 2016 to 2019. By using a combination of process mining techniques and traditional data mining, we were able to demonstrate that vaccinations in Victoria did not drop drastically–as other interactions did. On the contrary, we observed a surge of influenza and pneumococcus vaccinations in 2020, contradicting research findings of similar studies conducted in different geographical areas.
-
ItemUtility of Pan-Family Assays for Rapid Viral Screening: Reducing Delays in Public Health Responses During Pandemics( 2020)
Summary
Background
The SARS-CoV-2 pandemic has highlighted deficiencies in the testing capacity of many developed countries during the early stages of emerging pandemics. Here we describe the potential for pan-family viral assays to improve early accessibility of large-scale nucleic acid testing.Methods
Coronaviruses and SARS-CoV-2 were used as a case-study for investigating the utility of pan-family viral assays during the early stages of a novel pandemic. Specificity of a pan-coronavirus (Pan-CoV) assay for viral detection was assessed using the frequency of common human coronavirus (HCoV) species in key populations. A reported Pan-CoV assay was assessed to determine sensitivity to SARS-CoV-2 and 59 other coronavirus species. The resilience of the primer target regions of this assay to mutation was assessed in 8893 high quality SARS-CoV-2 genomes to predict ongoing utility during pandemic progression.Findings
Due to infection with common HCoV species, a Pan-CoV assay would return a false positive for as few as 1% of asymptomatic adults, but up to 30% of immunocompromised patients displaying symptoms of respiratory disease. Two of the four reported pan-coronavirus assays would have identified SARS-CoV-2 and we demonstrate that with small adjustments to the primers, these assays can accommodate novel variation observed in animal coronaviruses. The assay target region of one well established Pan-CoV assay is highly resistant to mutation compared to regions targeted by other widely applied SARS-CoV-2 RT-PCR assays.Interpretation
Pan-family assays have the potential to greatly assist management of emerging public health emergencies through prioritization of high-resolution testing or isolation measures, despite limitations in test specificity due to cross-reactivity with common pathogens. Targeting highly conserved genomic regions make pan-family assays robust and resilient to mutation of a given virus. This approach may be applicable to other viral families and has utility as part of a strategic stockpile of tests maintained to better contain spread of novel diseases prior to the widespread availability of specific assays. -
ItemBootstrapping Generalization of Process Models Discovered From Event Data( 2021-07-08)Process mining studies ways to derive value from process executions recorded in event logs of IT-systems, with process discovery the task of inferring a process model for an event log emitted by some unknown system. One quality criterion for discovered process models is generalization. Generalization seeks to quantify how well the discovered model describes future executions of the system, and is perhaps the least understood quality criterion in process mining. The lack of understanding is primarily a consequence of generalization seeking to measure properties over the entire future behavior of the system, when the only available sample of behavior is that provided by the event log itself. In this paper, we draw inspiration from computational statistics, and employ a bootstrap approach to estimate properties of a population based on a sample. Specifically, we define an estimator of the model’s generalization based on the event log it was discovered from, and then use bootstrapping to measure the generalization of the model with respect to the system, and its statistical significance. Experiments demonstrate the feasibility of the approach in industrial settings.