Monash University Research Portal
Not a member yet
41460 research outputs found
Sort by
The FLoRA Engine:Using Analytics to Measure and Facilitate Learners’ Own Regulation Activities
Get PDFThe focus of education is increasingly on learners’ ability to regulate their own learning within technology-enhanced learning environments. Prior research has shown that self-regulated learning (SRL) leads to better learning per-formance. However, many learners struggle to productively self-regulate their learning, as they typically need to navigate the myriad of cognitive, metacognitive, and motivational processes that SRL demands. To address these challenges, the FLoRA engine is developed to help students, workers, and professionals improve their SRL skills and become productive lifelong learners. FLoRA incorporates several learning tools that are grounded in SRL theory and enhanced with learning analytics (LA), aimed at improving learners’ mastery of different SRL skills. The engine tracks learners’ SRL behaviours during a learning task and provides automated scaffolding to help learners effectively regulate their learning. The main contributions of FLoRA include (1) creating instrumentation tools that unobtrusively collect intensively sampled, fine-grained, and temporally ordered trace data about learners’ learning actions; (2) building a trace parser that uses LA and related analytical techniques (e.g., process mining) to model and understand learners’ SRL processes; and (3) providing a scaffolding module that presents analytics-based adaptive, personalized scaffolds based on students’ learning progress. The architecture and implementation of the FLoRA engine are also discussed in this paper.</p
Audio Deepfake Detection:What Has Been Achieved and What Lies Ahead
Get PDFAdvancements in audio synthesis and manipulation technologies have reshaped applications such as personalised virtual assistants, voice cloning for creative content, and language learning tools. However, the misuse of these technologies to create audio deepfakes has raised serious concerns about security, privacy, and trust. Studies reveal that human judgement of deepfake audio is not always reliable, highlighting the urgent need for robust detection technologies to mitigate these risks. This paper provides a comprehensive survey of recent advancements in audio deepfake detection, with a focus on cutting-edge developments in the past few years. It begins by exploring the foundational methods of audio deepfake generation, including text-to-speech (TTS) and voice conversion (VC), followed by a review of datasets driving progress in the field. The survey then delves into detection approaches, covering frontend feature extraction, backend classification models, and end-to-end systems. Additionally, emerging topics such as privacy-preserving detection, explainability, and fairness are discussed. Finally, this paper identifies key challenges and outlines future directions for developing robust and scalable audio deepfake detection systems.</p
Enhancing Indoor Localization With Temporally-Aware Separable Group Shuffled CNNs and Skip Connections
No full textFingerprint-based indoor localization is the predominant localization approach for GPS-restricted environments due to its minimal hardware requirements. However, its performance is affected by signal fluctuations caused by shadowing, fading, and multipath effect, which necessitates models capable of capturing temporal variations. Numerous machine learning and deep learning algorithms have been introduced to surmount this limitation, within which the convolutional neural network (CNN) stands out as the most prominent. The 1D and 2D CNN models developed for indoor localization have the ability to extract spatial features but not temporal, leading to degradation in online localization accuracy. In contrast to 2D CNNs, 3D CNNs possess the ability to extract spatio-temporal information, but their computational complexity precludes real-time implementation. In this paper, a novel 3D-separable CNN for indoor localization is designed using skip connections and group shuffling. By employing depth-wise and point-wise convolutions, separable convolutions significantly reduce computational complexity, achieving a 10-fold improvement over conventional convolution, which facilitating real-time applications at the cost of reduced accuracy. Incorporating skip connections and group shuffling can help offset this accuracy decline. The 2D CNN, 2D separable CNN, and 3D CNN are used as benchmarks and evaluated on the UJILIB public dataset. Numerical results reveal that the proposed model is able to attain a positioning accuracy of 66.28% with an average positioning error of 1.04 meters in distance. Compared to the 2D separable CNN, the proposed 3D separable CNN outperforms it by achieving a 24.03% lower average positioning error.</p
Navigating fairness:practitioners’ understanding, challenges, and strategies in AI/ML development
No full textThe rise in the use of AI/ML applications across industries has sparked more discussions about the fairness of AI/ML in recent times. While prior research on the fairness of AI/ML exists, there is a lack of empirical studies focused on understanding perspectives and experiences of AI practitioners in developing a fair AI/ML system. Understanding AI practitioners’ perspectives and experiences on the fairness of AI/ML systems is important because they are directly involved in its development and deployment and their insights can offer valuable real-world perspectives on the challenges associated with ensuring fairness in AI/ML systems. We conducted semi-structured interviews with 22 AI practitioners to investigate their understanding of what a ‘fair AI/ML’ is, the challenges they face in developing a fair AI/ML system, the consequences of developing an unfair AI/ML system, and the strategies they employ to ensure AI/ML system fairness. By exploring AI practitioners’ perspectives and experiences, this study provides actionable insights to enhance AI/ML fairness, which may promote fairer systems, reduce bias, and foster public trust in AI technologies. Additionally, we also identify areas for further investigation and offer recommendations to aid AI practitioners and AI companies in navigating fairness.</p
Framing of innovation in urban Australian municipal climate policy
No full textDespite growing recognition of the importance of innovation in achieving net zero transitions, it remains unclear how to govern transformative innovation. Urban municipalities are consistently recognised for going above and beyond their state and national counterparts in their policy stances on climate change, yet there is limited investigation into how municipalities engage with and frame innovation in their decarbonisation policies, plans and strategies. In this study, we operationalise the three frames of innovation by Schot and Steinmueller (2018) to investigate the positionality of urban Australian municipalities in their innovation trajectory. Through a discourse analysis exploring 116 policy documents across 101 urban Australian municipalities, we find that despite ambitious rhetoric around innovation and system change, there is little evidence of transformative innovation within the policy deliverables. This has significant repercussions on the feasibility of locally driven net zero transitions and the consistent championing of municipalities within the literature. We call for greater attention to municipalities within the transformative innovation literature and propose a capability uplift in the sector is required to facilitate locally driven net zero transitions.</p
Serial Scammers and Attack of the Clones:How Scammers Coordinate Multiple Rug Pulls on Decentralized Exchanges
No full textWe explored the ubiquitous phenomenon of serial scammers, each of whom deployed dozens to thousands of addresses to conduct a series of similar Rug Pulls on popular decentralized exchanges. We first constructed two datasets of around 384,000 scammer addresses behind all one-day Simple Rug Pulls on Uniswap (Ethereum) and Pancakeswap (BSC), and identified distinctive scam patterns including star, chain, and major (scam-funding) flow. These patterns, which collectively cover about 40% of all scammer addresses in our datasets, reveal typical ways scammers run multiple Rug Pulls and organize the money flow among different addresses. We then studied the more general concept of scam cluster, which comprises scammer addresses linked together via direct ETH/BNB transfers or behind the same scam pools. We found that scam token contracts are highly similar within each cluster (average similarities > 70%) and dissimilar across different clusters (average similarities < 30%), corroborating our view that each cluster belongs to the same scammer/scam organization. Lastly, we analyze the scam profit of individual scam pools and clusters, employing a novel cluster-aware profit formula that takes into account the important role of wash traders. The analysis shows that the existing formula inflates the profit by at least 32% on Uniswap and 24% on Pancakeswap.</p
Finnish teacher students’ career choice motivations:a mixed methods study
Get PDFThis mixed-methods study analysed motivations to teach among Finnish first-year teacher students (N = 946) from early childhood, primary and special education programs, to discern potential additional motivations in a context where teaching is highly regarded and any differences between programs. Responses to an open-ended question revealed complementary motivational nuances to enrich the FIT-Choice scale in this context, such as ‘perceived social status’, highlighting the value of a mixed-methods approach. Highest-rated motivations on the FIT-Choice scale were social and intrinsic values, ability, and positive prior teaching and learning experiences; these were also most frequently mentioned of the FIT-Choice factors in open-ended responses, although correspondence at the individual level was modest. Minor differences occurred in career motivations between students from different programs.</p
Female neighbors and careers in science
Get PDFHow much does your neighbor impact your test scores and career? In this paper, we examine how an observable characteristic of same-age neighbors—their gender—affects a variety of high school and university outcomes. We exploit randomness in the gender composition of local cohorts at birth from one year to the next. In a setting in which school assignment is based on proximity to residential address, we define as neighbors all same-cohort peers who attend neighboring schools. Using new administrative data for the universe of students in consecutive cohorts in Greece, we find that a higher share of female neighbors improves both male and female students’ high school and university outcomes. We also find that female students are more likely to enroll in STEM disciplines that promote innovation and pursue more financially rewarding career paths when they are exposed to a higher share of female neighbors. We collect rich qualitative geographic data on communal spaces (e.g., churches, libraries, parks, Scouts and sports fields) to understand whether access to spaces of social interaction drives neighbor effects. We find that communal facilities amplify neighbor effects among females.</p
The safety actions of surfers in Victoria, Australia:assessing the impact of water safety and first aid training
No full textRecreational surfers in Victoria, Australia were surveyed every four weeks for 128 weeks about their beach safety actions when surfing. While previous studies have explored rescues performed by surfers over their surfing career/lifetime, this study is the first to provide data on safety action that incorporates a ‘per event’ denominator of exposure (safety action per 100 surfs). Victorian surfers reported giving safety advice, first aid and assistance in the water at average rates of 11.6, 0.4 and 1.1 times per 100 surfs respectively. We found that surfers with a board rescue qualification provided beach safety advice on average 3.1 more times per 100 surfs (29 % more often) and first aid 0.4 more times per 100 surfs than (or 3 times the rate of) surfers without this qualification. The difference between the rate of assistance provided by surfers with and without board rescue and first aid qualifications were not statistically significant. We estimate that surfers may provide safety advice up to 739,209 times, first aid 25,490 times and assistance 70,097 times per year in Victoria. The economic value of lives saved by survey respondents through surfer assistance in the water is estimated to be 3.8 billion (2023 AUD). Results confirm the importance of the safety actions of surfers to other beachgoers and point to the positive impact of board rescue training and qualifications on surfer beach safety actions.</p
Designing english curriculum courses for primary preservice teachers:a focus on the transformative potential of postmodern picture books
No full textIn this article, we document our experiences as teacher educators as we designed and implemented two courses that scaffold primary preservice teachers to engage critically with postmodern picture books and to explore a range of pedagogical practices for using postmodern picture books in classrooms with young children. Initially, our preservice teachers told us they did not have many experiences with postmodern picture books. Postmodern picture books are a special form of children’s literature that showcase some unique characteristics such as breaking boundaries, excess, indeterminacy and parody. In this article, our research investigation includes two case studies which draw on Schon’s classical approach to exploring the epistemology of our own practice through a reflective lens that brings together academic theory and professional practice. Firstly, we each recount our preservice teachers’ most adverse reactions to postmodern picture books. In response, we use the multiliteracies framework of the New London Group, that of situated practice, overt instruction, critical framing and transformed practice, to describe how we designed the learning activities and assessment tasks at two different universities in Australia. We do not attempt to generalise from our findings; rather, we explore the pedagogical framework that takes our preservice teachers from places of not knowing, resistance, and critique to one where they can articulate their understandings of postmodern picture books as social and cultural commentary and demonstrate a range of effective pedagogical applications.</p