26564 research outputs found
Sort by
Sentiment analysis of Nepali social media text with a hybrid deep learning model
No full textSentiment analysis (SA) has been a well-researched area in text mining, and signifcant progress has been made recently. Both traditional machine learning (ML) and recent deep learning (DL) methods have shown signifcant accomplishments in the SA for resource-rich languages. Various DL models, such as convolutional neural networks (CNNs), long short-term memory (LSTM), Bidirectional LSTM (Bi-LSTM), and pre-trained language models (PLMs), have been successfully used for SA in resource-rich settings. However, these models often struggle with low-resource languages like Nepali, primarily due to the limited availability of pre-trained models for handling complex language structures such as rich morphology, short texts, and unbalanced datasets. We propose a hybrid DL model that leverages contextual features from a PLM and spatial features from a convolutional module. For this, pre-trained multilingual embeddings from XLM-RoBERTa are utilised to comprehend the language context, which is then gated with a 1D convolution for local spatial pattern extraction. Finally, the softmax activation is applied with a dense layer to detect positive, negative, and neutral sentiments. We evaluate our model on three diverse Nepali SA datasets (D1, D2, and D3) and fnd the highest accuracy of 74.77% (D1), 79.52% (D2) and 55.82% (D3) compared to the state-of-the-art (SOTA) SA models for Nepali text.</p
Investigating the driving mechanism of megaproject social responsibility behavior: A system dynamics approach
No full textAlthough megaproject social responsibility (MSR) behavior has gained increasing attention from both academia and industry, it is still unclear how MSR behavior evolves over time. Utilizing system dynamics (SD) method, this study aims to investigate the driving mechanism behind MSR behavior from the perspective of a system evolution cycle. First, drawing on the Motivation-Opportunity-Ability (MOA) framework, the drivers of MSR behavior were inductively summarized through literature review and semi-structured interviews. Second, a causal loop diagram was employed to analyze the relationships between the drivers, leading to the development of a quantitative model. The valid empirical data from 162 respondents was collected to initialize parameters. Finally, system simulation was conducted using Vensim PLE. The findings reveals that the model effectively portrays the evolution of MSR behavior. The level of MSR behavior exhibits an overall growing trend, with the growth rate following an inverted U-shaped pattern over time. The simulation analysis underscores the pivotal roles of social responsibility culture, regulatory pressure and resource support. This study contributes to the body of the knowledge on megaproject management by offering a dynamic perspective on MSR behavior. The established SD model can be served as a practical tool and helps practitioners develop effective strategies for MSR governance.</p
How empowerment congruence impacts performance in cross-functional project teams: A polynomial regression analysis
No full textThis study deduces the empowerment alignment in project governance through examining the roles of (in)congruence between empowering leadership and psychological empowerment in informal leadership emergence, creativity, and performance outcomes. Using the data from 210 members from 69 cross-functional project teams, the results of polynomial regression and response surface analyses indicate that the high congruence between empowering leadership and perceived empowerment enhances informal leadership emergence, then promotes creativity to facilitate performance outcomes eventually. Informal leadership emerged under the insufficient empowerment condition with higher psychological empowerment rather than the excessive empowerment condition with higher empowering leadership. The study added to the project governance body of knowledge by confirming the promoting mechanism for project performance from the congruence perspective of leader-given and subordinate-perceived empowerment.</p
Nurse-led physical health interventions for people with mental illness: An integrative review of international literature
No full textBackground: People experiencing mental illness receive physical healthcare from nurses in a variety of settings including acute inpatient, secure extended care, forensic, and community services. While nurse-led clinical practice addressing sub-optimal consumer physical health is salient, a detailed understanding and description of the contribution by nurses to physical health interventions in people experiencing mental illness is not clearly articulated in the literature. Aims: The aim of this integrative review is to describe the state of knowledge on nurse-led physical health intervention for consumers, focusing on nursing roles, nursing assessment, and intervention settings. Methods: A systematic search of six databases using Medical Subject Headings from 2001 and 2022 inclusive was conducted. The Mixed Methods Appraisal Tool (MMAT) was utilised for quality appraisal. Results: Seventy-four studies were identified as “nurse-led”. Interventions were most common among community settings (n = 34, 46%). Nurses performed varied roles, often concurrently, including the collection of 341 physical health outcomes, and multiple roles with 225 distinct nursing actions identified across the included studies. A nurse as lead author was common among the included studies (n = 46, 62%). However, nurses were not always recognised for their efforts or contributions in authorship. Conclusions: There is potential gap in role recognition that should be considered when designing and reporting nurse-led physical health interventions.</p
Optimizing food waste anaerobic digestion in Kuwait: Experimental insights and empirical modelling using artificial neural networks
No full textThis study investigates, for the first time, the anaerobic digestion of food waste in Kuwait to optimize methane production through a combination of artificial neural network (ANN) modelling and continuous reactor experiments. The ANN model, utilizing eight hidden neurons and a 70-20-10 split for training, validation and testing sets, yielded mean squared error values of 0.0056, 0.0048 and 0.0059 and coefficient of determination (R²) values of 0.9942, 0.9986 and 0.9892, respectively. Methane percentages in biogas were predicted using six parameters: biomass type, pH, organic loading rate (OLR), hydraulic retention time (HRT), temperature and reactor volume. To validate the ANN results, continuous reactor experiments were conducted under an OLR of 3 kg VS m⁻³ d⁻¹ and HRT of 20 days at varying temperatures (35°C, 40°C, 45°C, 50°C and 55°C). The experiments demonstrated optimal methane production in the mesophilic range, with ANN predictions closely aligning with experimental data up to 45°C. However, deviations were observed at higher temperatures, particularly under thermophilic conditions beyond 50°C. This study provides novel insights into waste-to-energy initiatives in Kuwait and highlights the potential of integrating computational models with empirical data to enhance biogas production processes.</p
DMP_1466 A woman's place is at the chalkface: using autoethnography alongside allegory to inform historical fiction writing
No full textThis qualitative practice-led research project resulted in a historical fiction novel creative artefact and an exegesis that investigated the writing practice and final creation through a neurodivergent lens (ADHD). As creative practice-led research is a field that is growing and includes a variety of experimental designs, this research project aimed to fill the current gap in the literature with a hybrid approach that incorporated autoethnography, allegory and extended metaphor, Todorov's narrative theory of equilibrium and historical (archival) research methods. The contextual background derived from my experiences of teaching in a high-school between March and August 2020 after the COVID-19 pandemic arrived in Australia. My research design made a comparison of female and governmental officials' responses during the 'Spanish' pneumonic influenza epidemic of 1918-1919 and the COVID-19 pandemic of 2020. Both health disasters disrupted societal functioning and traces of policies that historically disadvantaged women through an expectation they would shoulder the burden of caring for their communities can still be found today. Autoethnographic writing practices have been shown to provide therapeutic benefits to the writer and validation for the reader as a means to compare their experiences. Conveying these findings through an allegorical creative artefact in the form of a novel and an exegesis, I aimed to share understanding not only with other female teachers and women late-diagnosed with ADHD, but also individuals who seek an opportunity to creatively and critically express themselves during or after exceptional situations when they are bound by policies that require confidentiality.</p
Sustainable and Adaptive Artificial Intelligence-Powered Internet of Things Solution for Scalable Smart Irrigation Management
No full textWater breathes life into our parks by transforming them into lush and vibrant expanses of green. With the surge in population growth and the ever-present spectre of climatic changes, the urgency to implement effective water conservation measures in irrigation practices has intensified. Balancing optimal irrigation with avoiding excessive watering poses a critical challenge. Over-irrigation and inadequate water management within parks and gardens exacerbate the wastage of water resources, potentially leading to seepage, runoff, and the leaching of vital nutrients into nearby streams. In stark contrast, under-irrigation yields diminished plant growth and an unattractive aesthetic. A significant motivator for improving the system was addressing public complaints about waterlogging and the park's overall appearance. Thus, irrigation management becomes paramount in maintaining the verdant and inviting landscape we cherish. Complicating matters, variability in soil properties, environmental conditions, and landscape features can create a patchwork of soil moisture levels. This non-uniform distribution within a single landscape necessitates a comprehensive understanding of soil moisture distribution. However, existing irrigation systems often struggle to adapt to such variability due to limitations in sensor placement, underutilisation of local environmental data, and the absence of predictive modelling techniques. The evaluation of the existing irrigation system at Esplanade Park revealed inefficiencies in both water usage and resource management. Station 11 and Station 12 operated daily for 15 and 25 minutes respectively, resulting in a combined water usage of 2,160 litres per m² per year. Leachate collectors installed below the root zone recorded significant water percolation, indicating that water was draining beyond the root depth and being wasted. The system lacked predictive or adaptive control, leading to over-irrigation and excessive energy use. These findings highlight the urgent need for a smarter, data-driven irrigation approach. To overcome these challenges, this study explored an intelligent and innovative alternative solution to address this complexity. Based on extensive experiments and study, this thesis presents a smart and precise irrigation system comprised of various intelligent techniques and automation using machine learning and Internet of Things (IoT) technologies. To collect initial data, this study started with a dual-Electromagnetic scan of the soil, spanning a period, to unveil the intricate distribution of soil moisture across an experimental site. Based on this data, soil moisture is stratified into three distinct zones – wet, semi-wet, and dry. This categorisation lays the groundwork for the strategic placement of soil moisture sensors. An advanced open-source, cloud-based IoT system has been meticulously crafted to introduce a new era of precision. This state-of-the-art system is designed to capture real-time soil moisture and weather data, processing this wealth of information to make informed decisions regarding park irrigation needs. Using real-time data, this study has explored various Artificial Intelligence (AI) and machine learning-based models to forecast soil moisture content. The system uses incremental machine learning to learn from the shifting environmental context. One of the pivotal challenges faced by smart irrigation technology is its adaptability, which is often constrained not only by the large number of soil moisture sensors required for real-time monitoring, but also by the complexity of determining optimal sensor placement across varying spatial and depth profiles, both of which significantly increase installation and maintenance costs. A machine learning-based solution is unveiled to combat this challenge head-on, enabling the precise vertical and horizontal soil moisture assessment with a reduced sensor count. A three-dimensional recursive model has been developed to address this challenge, utilizing nearest-neighbour interpolation, linear interpolation, and cubic interpolation machine learning techniques. This model combines real-time soil moisture and weather data with on-field metrics like infiltration rate, pH, and ET to predict soil moisture at varying depths (7 cm, 30 cm, and 90 cm) with an accuracy of 96 %. In comparison, the linear three-dimensional recursive interpolation achieved an accuracy of 94 % at a depth of 120 cm. With the help of real-time data from the field using an IoT system, on-field metrics like infiltration rate, pH, and AI-driven predicted soil moisture; this study has innovatively developed an intelligent irrigation system with a remarkable accuracy rate of 94 % in predicting the park's irrigation needs. I have used statistical and deep machine-learning models to develop an intelligent, innovative irrigation system that is precise, adaptable, and sustainable. Furthermore, to predict soil moisture in the future, the developed intelligent irrigation system uses deep learning models like Long Short-Term Memory (LSTM) to forecast moisture with 95 % accuracy. During the testing period, the system enabled sprinkler shutdown for nine days in a month, resulting in an estimated water saving of 648 litres per m² per year. In conclusion, the confluence of cutting-edge technology, environmental consciousness, and innovative machine-learning methodologies propels us toward an era of sustainable and precise smart irrigation. This reduction in sensor numbers strengthens the adaptability of smart irrigation technology in agriculture and park management but also minimises water usage, limits chemical seepage into nearby streams, elevates crop yields, and reduces maintenance costs. The findings support the potential cost-efficiency and operational benefits of adopting this system for large-scale park irrigation. This research has put forward a transformative water resources management system for parks, ushering in a brighter, greener future for our parks and agricultural landscapes.</p
Assessing the development of statin associated myopathy in an old obese rat model and the efficacy of a novel pharmacological intervention to reduce myopathy development.
No full textBACKGROUND Lifestyle choices such as inactivity and diets high in saturated fats has led to an increase in mortality associated with cardiovascular disease (CVD) globally. The main strategy to reduce CVD risk has been targeting elevations in cholesterol with a range of medications. Of these medications, statins, or 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors are the most effective and most prescribed. Statins reduce low density lipoprotein concentrations (bad cholesterol) in circulation and formation within hepatic cells. Although statins are proficient at reducing CVD risk, side effects can develop and the most common is myopathy. Myopathy can range from muscle aches and pain, to significant breakdown of skeletal muscle and is reported to affect 100 of every 1 000 000 statin patients. Presently the mechanisms behind statin associated myopathy (SAM) are unknown and studies are underway to determine this, however what is unknown, is how age and obesity impact upon SAM development. This study aims to investigate how aging and obesity impact upon the development of SAM and whether a co-therapy treatment using L-arginine can reduce the onset and/or severity of SAM development in an aged biological female, obese rodent model. METHODS In addition to obtaining biometric data throughout the treatment period, terminal assessments of electrical and mechanical function of the heart were undertaken, along with organ bath experiments examining the reactivity of small and large blood vessels, and skeletal muscle function. Blood and tissue samples were also collected for biochemical and genetic analysis, qPCR, to identify possible mechanisms underlying SAM development. RESULTS Functionally, skeletal muscle stimulation response for statin treated animals indicated an elongation in fused tetanus and muscle mass reductions compared to age matched control animals. Molecular assessment indicated an increase in gene expression implicated in the control of oxidative stress and reduced glucose metabolism, with both pathways impacting mitochondrial function. L-arginine treatment in biological female aged and obese animals showed a sustained and improved function and response of vascular smooth muscles both with and without statin treatment. CONCLUSION These findings suggests that high dose statin treatment can negatively impact skeletal muscle function, often in concert with obesity and age. This study utilised an animal model that closely represents the population most likely prescribed statins to assess skeletal muscle, vascular smooth muscle and cardiac muscle health and function to provide greater detail on the current known literature of high dose statin treatment.</p
Sleep timing across the lifespan of Australian adults
No full textThe aim of this study was to examine sleep timing across the lifespan of Australian adults. A cross-sectional design was used to collect information on subjective sleep timing from 1225 participants (52.3% female) during a telephone interview. The participants were aged from 18 to over 80 and were grouped according to their age using 10-year increments (e.g., 18–29 y, 30–39 y, etc.). There was a diverse distribution across the lifespans, with the largest proportion of participants being from the 60–69 age group (22.8%). Participants were predominantly from New South Wales, Queensland, and Victoria. Younger adults reported going to bed later (p </p
Reflecting on ‘doing better’: Inclusive and accessible qualitative research methods
No full textWhen we do not include, or account for difference, in qualitative research studies, outcomes create insufficient data that limits innovations in practice, policy or legislation development. Accessibility and inclusion in qualitative research should accommodate all narratives, thus challenging the creation of epistemic and social injustice.
The goal of this special issue came about in response to Watharow and Wayland’s paper on inclusion of d/Deafblind participants in research (2022) generating conversations on intersectionality and representation of all people in qualitative research. The special issue is framed through Fricker’s, 2007 Epistemic Injustice: Power and the Ethics of Knowing. Acknowledging that true knowledge building can only occur if research is designed in socially just ways. Knowledge building must seek the standpoints and testimonies of the people and communities the knowledge seeks to serve.</p