69238 research outputs found
Sort by
Learning Māori beyond the classroom with the linguistic landscape: a socio-cognitive perspective
No full textThis study explores the use of the linguistic landscape (LL) by two
recent migrants to New Zealand as a resource for language
learning beyond the classroom (LBC) from a socio-cognitive
perspective. With a focus on learning Māori, we inquire into these
beginner-level language learners’ experience of using bilingual
public signage as the primary learning resource over a fourmonth
period, with a specific focus on the learners’ interest,
perceived self-efficacy, and time investment during the LBC
experience. Using a purpose-built dataset of bilingual signage in
Auckland, the learners identified criteria to categorize individual
signs according to difficulty level, and they tracked their learning
experience using monthly questionnaires (interest and selfefficacy),
a time sheet, and personal journals. The learners
displayed individual differences in their approaches to LBC, and
in their perceived self-efficacy. We discuss these differences, using
the journals to provide insight into learners’ thoughts, actions
and emotions. We provide recommendations for autonomous
language learning with the LL, and for the promotion of bilingual
public signage as a tool for language revitalization
ALEA: Active Learning-Enhanced Multimodal Entity Alignment with Semantic Modality Imputation
No full textIntegrating multimodal knowledge graphs, referred to (MMKGs), from heterogeneous sources is essential for comprehensive knowledge representation and reasoning. Nonetheless, existing multimodal entity alignment (MMEA) methods struggle with missing modalities and semantic inconsistencies between disparate sources. These limitations compromise alignment robustness, especially in low-resource scenarios with limited seed pairs.
To bridge this gap, we propose ALEA, a novel MMEA framework that integrates semantic calibration and active learning to improve alignment accuracy and robustness. ALEA consists of three key components: Latent Semantic Learning (LSL) synthesizes embeddings for missing modalities to preserve the distributional semantics of entities; Latent Semantic Calibration (LSC) refines entity representations to mitigate semantic inconsistencies across MMKGs; and Active Candidate Selection (ACS) dynamically selects high-confidence and diverse candidate pairs within a learnable budget to enhance low-resource settings alignment.
Extensive experiments on benchmark datasets demonstrate that ALEA consistently outperforms state-of-the-art baselines, improving MRR by 2.02% and 2.85% over the best existing methods. Analytical ablation studies further validate ALEA’s ability to bridge semantic gaps and handle incomplete MMKGs. Additionally, ALEA significantly enhances alignment robustness in low-resource scenarios, improving accuracy by nearly 4.7% over the non-active learning variant. These results highlight ALEA’s potential as an effective solution for scalable and robust multimodal entity alignment in real-world MMKG integration
Exploring and Optimising DNA Collection and Extraction from Forensic Samples
No full textDNA analysis is an important aspect of forensic investigations as it helps link individuals to a crime. Forensic scientists face many challenges when attempting to extract DNA from crime scene samples. Samples often contain small amounts of DNA, and analysis can be further challenged by other factors. Examples include exposure to heat, metal surfaces and fingerprint treatments all that may interfere with DNA analyses.
This research focused on the optimisation of the sampling and DNA extraction steps. The objective was to determine the best materials and techniques for sampling and extracting the DNA from forensic samples.
In the first phase swabs (Cultiplast®, SceneSafeTM, Sarstedt, and DNAMICTM) were tested with different wetting agents (Water, Ethanol, BSA, and EDTA) on body fluid samples that had been placed on acetate sheets. It was found that the performance of different sampling materials differed based on the body fluid being sampled. The interpretation of the results was affected by the small sample size and variation. It was found that different combinations of sampling methods were relatively more successful depending on the body fluid type. The use of BSA and DNAMICTM swabs were a promising combination for blood sample collection. This study also showed that Cultiplast® swabs, SceneSafeTM swabs coupled with water and ethanol are fit for purpose, while EDTA and Sarstedt swabs gave relatively less successful results.
The next stage focused on testing different extraction methods (DNA IQTM and the QIAamp DNA Investigator Kit) on different challenging samples: fired cartridge cases, heated metals, and treated fingerprint samples. The DNA Investigator kit performed better than DNA IQ™ for fired cartridge cases and heated metals. DNA IQ™ seemed to be less affected by the presence of fingerprint powder and performed better for this sample type.
Low level DNA results were obtained for the three challenging sample types. DNA profiles suitable for comparison purposes were not obtained and this was not entirely unexpected. Useful insights could still be obtained into combinations of sampling methods and DNA extraction methods to form a starting point for further study
Comparison of antimicrobial efficacy and safety of pulsed versus continuous wave UVC
No full textPURPOSE: Preclinical studies have confirmed efficacy of ultraviolet C (UVC) light in managing superficial corneal infections. This study sought to establish if delivering UVC in pulsed form enhances antimicrobial efficacy compared to continuous delivery, and whether pulsed delivery in ocular tissue results in deeper penetration or introduces additional safety concerns. METHODS: This study compared antimicrobial efficacy, depth of penetration, and safety of continuous versus pulsed wave delivery of UVC (20 Hz, 50% duty cycle) in three experimental setups. Firstly, efficacy was assessed using a simulated in vitro corneal wound model infected with bioluminescent P. aeruginosa, comparing matched-fluence, 0 to 120 s continuous wave versus 0 to 240 s pulsed wave exposures. Secondly, penetrability was evaluated in an ex vivo porcine corneal model (0 to 650 µm thickness). Lastly, safety was analyzed by immunohistochemistry to assess DNA photoproducts, cyclobutene pyrimidine dimers (CPD), focusing on their spatial distribution and density after UVC exposure (579 mJ/cm2). RESULTS: Comparable antimicrobial efficacy was observed for continuous and pulsed wave UVC (50% duty cycle) for all exposure durations (p > 0.05), except the 40 s pulsed wave, which was more effective than the 20 s continuous wave (p 0.05). Immunohistochemistry confirmed CPD were confined to the superficial corneal epithelial layers, with no significant differences in depth or extent of CPD formation between pulsed and continuous wave delivery modes (p > 0.05). CONCLUSIONS: In in vitro and ex vivo testing, pulsed wave UVC demonstrated antimicrobial efficacy that was at least as good as continuous wave delivery, demonstrated comparable corneal depth penetration, and similar spatial distribution of CPD
The Evolution and Enactment of CT Policy Initiatives in the Chinese Education Context
No full textThis thesis investigates the evolution and enactment of critical thinking (CT) policies within China’s secondary education system, framed through Bowe et al.’s (1992) policy trajectory model. Since 1999, the Chinese government has promoted CT-related elements as part of quality-oriented education reform; however, substantial curriculum integration did not occur until the revised curricula in 2017 and 2020. This research analyses representations of CT within key policy documents (Study 1) and explores how schools enact CT initiatives in practice (Study 2), thereby highlighting alignments and disconnections within Chinese educational reform.
Study 1 employs Foucault’s genealogical discourse analysis to examine the context of policy text production. It analyses four national policies and three curriculum documents to understand how CT has been constructed over time. Study 2 shifts to the context of practice, using Ball et al.’s (2011) policy enactment theory to explore CT initiatives in two senior high schools in different regions of China. A case study methodology, drawing on interviews, observations, and policy documents, captures the dynamics between institutional contexts and policy actors.
Five key findings emerged, revealing a complex CT policy landscape. First, national policy shifted notably in 2016 with the Core competencies and values for Chinese Students, advocating four CT dimensions: skills, dispositions, cognition, and ethics. This policy led to CT’s diversified inclusion in the 2017 curriculum. However, a disconnect persists: national policies emphasise CT’s ethical dimensions, while curricula focus on skills. At the school level, policy enactment varies significantly. One school established a systematic mechanism for CT integration through diverse policy roles and external partnerships, while the other exhibited fragmented practices due to the absence of structured enactment mechanisms, despite students’ strong interest in CT. Challenges to policy enactment include limited support from local education bureaucrats and tensions between collective and individual orientations towards CT.
Overall, this thesis reveals recontextualised CT knowledge in official discourses and the nuanced dynamics of policy enactment in senior high schools. It therefore offers a valuable perspective on how CT is constructed and circulated within Chinese education contexts
A Multi-Chamber Chewing Robot for In Vitro Studies of Food Texture Changes during Mastication: Design, Control and Experiments
No full textMastication is defined as the process where the food is placed into the oral cavity and being crushed and grounded by rhythmic teeth movements assisted by the tongue and cheek muscles, softened by the secreted saliva and eventually forming into a bolus which is ready to be swallowed. Understanding the changes in food texture during mastication and the interaction between food and the masticatory system is critical for the development of food products and dental implants. In vivo studies of food texture changes are often constrained by ethical considerations, limited access, and challenges in ensuring repetitive and objective observations. As a result, in vitro studies using chewing robots have gained prominence. However, most existing chewing robots can only replicate limited aspects of oral functions, and the efficiency of in vitro studies is limited. Therefore, a new generation of chewing robot was conceptualized and developed to mimic the human mastication process and was applied to study the effects of oral condition parameters on the food texture changes during the chewing process.
By analysing the pros and cons of existing chewing robots, the technical requirements for the new design were developed. A new cam-follower based chewing robot with three oral chamber is proposed to improve the efficiency of in vitro experiments. Cams with different profiles were installed on a camshaft and be actuated to generate different chewing trajectories simultaneously. A remote centre compliance (RCC) mechanism is installed between the upper molar and follower to limit the maximum chewing force applied to the robot, as well as automatically align molars when the molars were misaligned. Additionally, a V-shaped food repositioning mechanism, requiring no additional actuation, is integrated into the oral chambers to facilitate food manipulation, which is validated by experiments. A peristaltic pump is installed to add artificial saliva to the oral chamber during the chewing process. Besides, load cells are installed beneath the oral chambers to measure the real-time chewing resistances.
This thesis can be divided into two parts, where the first one discusses the conceptualization and design of the robot system, kinematics and dynamics. The mechatronics system including the actuation and sensing of the robot and a graphic user interface (GUI) is illustrated. The second part centres on measuring real human chewing motions when chewing various foods and applying them as references for the chewing robot. By applying position control of the motor movement to enhance the precision of the chewing motion, the robot can mimic the natural variability found in human mastication.
Robotic chewing experiments were conducted to analyse the food texture changes during the masticatory process. Roasted peanuts and white bread were applied as sample foods. The number of chews required to chew different foods until they were ready to be swallowed was recorded through in vivo experiments. Different chewing trajectories of molar, chewing forces and flow rate of artificial saliva were applied and combined to simulate different oral processing conditions in the chewing robot. The food samples were chewed by the robot for 0, 25, 50, 75 and 100% of the recorded number of safe swallowing chews. Nine food texture properties were measured via texture profile analysis (TPA) by using texture analyser. Principal components analysis (PCA) and partial least squares regression (PLSR) were applied to study the correlation between the oral processing conditions and food texture properties. The results indicated that for roasted peanuts, hardness, adhesive force, and cohesiveness exhibited strong correlations with chewing cycles, suggesting a close linear relationship between these variables and the number of chews. In contrast, for white bread, the correlations of these variables with chewing cycles were less pronounced. Finally, the mechanisms of food texture changes during robotic chewing were analysed and explained. For roasted peanuts, the changes in hardness, adhesive force, and cohesiveness were primarily influenced by the stages of chewing, whereas springiness and the springiness index were mainly dependent on saliva secretion rate, indicating a higher chewing efficiency of adolescents and young people than children and older people. In the case of white bread, hardness was predominantly determined by occlusal force during the early stage of chewing and by saliva secretion rate in the later stage. Adhesive force was mainly influenced by saliva secretion rate, while cohesiveness was largely affected by chewing trajectory. The gumminess, chewiness, and chewiness index of white bread were primarily affected by saliva secretion rate throughout the chewing process, indicating a higher chewing efficiency of young people than older people when chewing resilient foods.
The development of the chewing robot effectively explains the food texture changes during the chewing process with human mastication being mimicked. The chewing experiments on roasted peanuts and white bread proves the robot as a suitable tool for carry out chewing simulations on foods. It is validated that the effect of chewing conditions of different groups of people on food texture changes can be studied through in vitro methods
Can realists say that state wrongdoing corrupts legitimacy?
No full textLiberal political theory typically holds that injustice can corrupt legitimacy. Realists critique such views as ‘moralist,’ arguing that justice and legitimacy are not the same thing. For realists, even profoundly unjust states can satisfy basic legitimation criteria. While realism may appear hostile to the thesis that injustice corrupts legitimacy, this paper demonstrates that realist theory is more hospitable to that claim than one might expect
Impact of Māori- and Pacific-Specific SLC22A3 Genetic Variants on the Efficacy of Platinum-Based Chemotherapy Agents
No full textAim/Hypothesis: The intracellular uptake of platinum-based chemotherapeutic agents, particularly oxaliplatin, is mediated by multiple different transporters, including organic cation transporters (OCTs). Organic Cation Transporter 3 (OCT3), which is encoded by the SLC22A3 gene. Genetic variants in SLC22A3 gene have shown impact on OCT3 mediated platinum drug transport. We identified two variants in SLC22A3 gene that are unique to Māori and Pacific populations, the impact of which on platinum drug transport haven’t been studied yet. Therefore, this study aims to determine the impact of these two variants on the effect of platinum-based chemotherapeutic agents. As SLC22A3 transports a wide range of drug substrates, the second part of the study aims to examine the impact of the presence of other substrate drugs on the effectiveness of the platinum-based drugs in colorectal cancer cells (CRC).
Methods: Cell viability assays were used to assess the cytotoxicity of platinum drugs in SLC22A3 overexpressing HEK293 cells and colorectal cancer cells with endogenous SLC22A3 expression, including drug combinations to evaluate the impact of other substrates (in CRC cells). Colony formation and wound healing assays were used to examine the effects of oxaliplatin alone and in combination with other drugs on cancer cell growth and migration. Finally, RT-qPCR was performed to analyse transporter gene expression in all HEK293 cell lines and colorectal cancer cell lines that used in the study.
Results: SLC22A3 variants T44M and V494M overexpression HEK293 cells exhibited increased sensitivity to oxaliplatin, indicating that these variants enhance oxaliplatin uptake and thereby increase its cytotoxic effects. Cisplatin displayed similar effects on cell viability in both wildtype and the two variant expressing cell lines. Additionally, among the transporters examined in this study, OCT2 mRNA expression level was higher than OCT3 in the colorectal cancer cell lines, suggesting a potential dominant role for OCT2 in mediating oxaliplatin uptake in these cells. Propranolol, an β-adrenergic blocker and known substrate of organic cation transporters inhibits cancer cell growth when used alone. When combined with oxaliplatin, cytotoxicity effect were further increased in CRC cells. Isoproterenol, a β-adrenergic agonist, showed minimal impact on oxaliplatin-induced cytotoxicity; while corticosterone, an organic cation transporter inhibitor, reduced oxaliplatin’s cytotoxicity in certain cell lines when used in combination treatment.
Conclusion/interpretation: This study suggests that SLC22A3 variants (T44M and V494M) enhance oxaliplatin uptake, highlighting their potential to either cause increased cytotoxicity or increased anti-tumour efficacy. Additionally, colorectal cancer cells that were sensitive to oxaliplatin and propranolol also exhibited high SLC22A2 expression, suggesting a possible association that warrants further experimental validation
The impact of area deprivation on access to and utilization of health services in the last year of life: A retrospective population-based cohort study
No full textBackground: The healthcare needs of people living in areas of high deprivation are complicated by the cumulative effect
of the sociodemographic factors known to impact on health outcomes, such as income, housing and education. Of note,
for people living in more deprived areas, life expectancy is shorter and the onset of chronic disease and multimorbidity
occurs much earlier. While the relationship between area deprivation and access to palliative care is becoming more
widely researched, the vast majority of studies to date have focused on referrals to specialist palliative care services. This
is problematic given the dominant model of generalist–specialist palliative care in high-income countries which assumes
that most people will have a level of palliative care need that can be managed by non-specialist palliative care services.
Objective: To identify associations between area deprivation and the use of generalist and specialist palliative care
services in the last year of life.
Design: A retrospective population-based cohort study.
Methods: People aged over 18 years who died between January 2015 and December 2020 were identified within one
geographical area of Aotearoa New Zealand. Using the National Health Identifier, deaths were matched to generalist
and specialist palliative care service data.
Results: A significant association was found between area deprivation and health service use in the last year of life. Of
note, people living in rural areas of deprivation were significantly less likely to receive a hospital (p=<0.000) or inpatient
hospice admission (p=<0.000). They were also less likely to have contact with their general practitioner (p=0.007) or
experience a specialist outpatient clinic appointment (p=0.001).
Conclusion: This study has revealed inequities in health service use across generalist and specialist palliative care
services for people living in areas of deprivation. Of note, findings have highlighted how rurality amplifies inequities
in access to appropriate palliative care. Further research is needed to better understand the consequences of these
apparent inequities
Impacts of a community-wide infrastructural intervention on children's active school travel: A controlled before-after study
No full textBackground Active school travel (AST) is beneficial for children's physical, social, and mental well-being. However, AST is low globally, and Aotearoa New Zealand has one of the lowest rates of AST worldwide. Te Ara Mua – Future Streets (TAMFS) was a randomised controlled before-after study of neighbourhood-wide infrastructural changes to make it easier for residents to get around actively. The intervention was conducted in Ngā Hau Māngere, in Aotearoa New Zealand. This study examines the impact of TAMFS on children's AST over time. Methods Data for the current study were gathered from resident surveys, conducted at baseline (2014), early follow-up (2017), and long-term follow-ups (2021, 2023). The primary outcome for this study was the proportion of children using AST, analysed using mixed model logistic regression and difference-in-differences (DID) analysis. Coding of open-ended responses to the main reason for children's travel mode to school was also undertaken. Results AST rates declined in both intervention and control areas over time, with a significant drop in the intervention area immediately post-intervention (2017). By 2023, the difference between areas was minimal. Living close to school was the main reason for getting to school actively, while a mix of distance (longer), convenience, and safety concerns were key decisionmakers for parents of passive travellers. Conclusion While environmental changes are important, future interventions would benefit from considering broader contextual factors and integrating community engagement and programming to address barriers to AST and support shifts to active transport modes