110830 research outputs found
Sort by
Corpus Linguistics and Language Testing and Assessment
No full textCorpus linguistics is a scientific method for the theoretical and applied analysis of language. It allows the investigation of large digital collections of texts (written texts and transcripts of spoken language) and provides empirical evidence for the fine-grained analysis of linguistic features. One of the applications of corpus methods is in the field of language testing and assessment. Corpus linguistics can be beneficial to language testing and assessment informing the test cycle at different stages, for test design and validation, examiner training and standardization, as well as for developing automated scoring and feedback systems in computer-based tests
Investigating molecular pathways and identifying genes and microbes that promote heat stress resistance and healthy ageing
Get PDFSeveral studies have linked stress resistance to lifespan extension, and ageing has been associated with dysbiosis. With C. elegans longevity correlating with severe heat resistance and the recent recognition of a key role for the gut microbiota in defining worm healthspan, I looked at how C. elegans gut microbes found in its natural environment may interact with the worm physiology to mediate stress resistance and healthy ageing. I exposed young adult C. elegans to their natural gut microbiota and assayed for heat (42°C) and oxidative stress (7% tertbutyl hydroperoxide) resistance using LFASS (Label-Free Automated Survival Scoring) assays. To determine if microbial effects on stress resistance involve known gut-brain axis pathways, I conducted LFASS assays on Insulin/IGF-1 Signalling (IIS) and Kynurenine Pathway (KP) C. elegans mutants. My results showed that both the KP and IIS pathways are involved in severe stress resistance, with bacterial strains modulating the KP differentially and the gut microbiota influencing stress responses through both IIS-dependent and independent mechanisms. I then conducted health-span (movement, brood size) experiments and lifespan assays to determine whether enhanced stress resistance translates into healthier ageing and longevity. MYb71 (Ochrobactrum vermis), MYb330 (Pseudomonas), BIGb170 (Sphingobacterium multivorum), MYb396(a) (Comamonas), and MYb21 (Comamonas), were found to be beneficial, as they improved stress resistance without negatively impacting reproduction. Enhanced stress resistance did not lead to increased lifespan, suggesting that stress resistance and longevity are physiologically distinct. To identify new host genetic pathways potentially mediating gut microbiotas effects on severe stress resistance and healthy ageing, I conducted a genome-wide RNAi screen on young adult C. elegans exposed to severe heat stress. I found known thermoprotective genes, F44E5.5, sip-1, and dnj-5, and thermosensitising gene skn-1. I also identified some new thermoprotective candidate genes, such as clec-20, F20G2.5, pals-30 (immunity) and R08F11.1, fat-7 (fat metabolism). The genes from the RNAi screen will be used in conjunction with the beneficial microbial isolates to determine whether they produce a synergistic response
Neurolinguistic signatures of microvariation and language change : Individual production variability predicts morphosyntactic processing differences
No full textThis study investigates the neurophysiological and behavioral effects of exposure to grammatical microvariation, focusing on the relationship between individual production patterns and on-line sentence processing. We examine a unique context of ongoing morphosyntactic change in Norwegian participial agreement, which gives rise to widespread, near-continuous variation in agreement marking across dialects and speakers. Event-related potentials (ERPs) were recorded from 156 participants drawn from two dialect regions with extensive inter- and intra-speaker variability, as they read sentences in standardized Norwegian Bokmål containing either invariant (control) or variable/changing (target) grammatical forms. Results show that both group-level dialect background and individual production frequencies modulate sentence processing. Participants from regions undergoing dialect change exhibit attenuated P600 effects and more gradient acceptability judgments, and neural sensitivity to participial agreement violations is continuously modulated by individual age and spoken agreement rates. These findings provide evidence that language processing is dynamically shaped by fine-grained individual differences in linguistic experience – including community-level exposure, personal production frequencies, and sociolinguistic positioning within ongoing grammatical change – and that even narrow between- and within-individual variation in dialectal grammar influences integration costs during standard language comprehension
Assessing the drivers of non-communicable diseases prevention activities in primary health care facilities in Ghana : a case study of some selected districts
No full textBackground: Primary health care (PHC) is pivotal to equity and disease control in low- and middle-income countries. In Ghana, the rising burden of non-communicable diseases (NCDs) demands a shift from curative to preventive PHC. Evidence on the extent and drivers of PHC-led prevention, however, remains limited. Methods: A cross‑sectional, facility‑based survey was conducted between November 2023 and March 2024, covering 210 primary‑health‑care facilities in 10 randomly selected districts of Greater Accra and the Eastern Region, Ghana. An adapted WHO‑SARA questionnaire was administered to facility heads or senior clinicians, capturing facility type, ownership, service‑readiness scores, and four self‑reported NCD‑prevention activities, community durbars, home visits, screening tests, and nutritional counselling. Multivariable logistic regression identified predictors of each activity, and a zero‑inflated negative‑binomial model evaluated the association between preventive engagement and NCD‑related outpatient‑department visits. Results: Overall, 31% of facilities conducted Community Durbars, 63% Home Visits, 59% Screening Tests and 54% Nutritional Counselling. CHPS compounds (56% of facilities) were the most active—81% provided home visits, while only 8% of private clinics did so. CHPS status strongly predicted home visits (OR 26.89, 95% CI 1.20–604.85). Facilities offering Community Durbars (β 1.24, p < 0.01) or Nutritional Counselling (β 1.72, p < 0.01) recorded higher NCD-related OPD use. Higher district NHIS coverage was inversely associated with preventive engagement (OR 0.26, p < 0.01). Conclusion: NCD prevention in Ghana is led by lower-tier, government PHC facilities, yet resource gaps and treatment-centred financing limit wider uptake. Re-aligning NHIS incentives, investing in CHPS infrastructure and tailoring outreach to underserved groups especially men are critical for shifting Ghana’s PHC from reactive care to sustainable NCD prevention
Exact constrained-training neural networks for confidential 8-bit arithmetic primitives in code obfuscation
Get PDFNeural networks are rarely utilized for exact arithmetic computations. However, we aim to leverage them as accurate computing units in code obfuscation scenarios. The effectiveness of traditional code obfuscation methods is increasingly compromised by advanced reverse-engineering tools, which exploit the semantic transparency of arithmetic operations. To address this, we propose constrained-training neural networks tailored to 8-bit integer addition and multiplication, the most common operations in security-critical software. For addition, we introduce a constrained training algorithm that integrates weight clipping, linearity-enforcing loss terms, and boundary-case oversampling, enabling convergence to 100% accuracy across the entire domain. For multiplication, we design the adaptive symbol-gated NALU (ASG-NALU), an improved 4-bit multiplier that achieves exact results with reduced complexity. Combined with a cascade decomposition strategy, it extends to 8-bit multiplication with guaranteed correctness. Experiments confirm 100% in-domain accuracy, while out-of-domain inputs trigger catastrophic failures that act as natural traps, providing hidden security checks against dynamic analysis. These results establish exact constrained-training neural networks as confidential arithmetic primitives and firmly position neural arithmetic as a promising approach for advancing code obfuscation techniques
Tuning crystal field stabilization energy through configurational entropy design in high-entropy oxide sodium-ion battery cathodes
Get PDFHigh entropy layered oxides have emerged as promising cathode materials for sodium-ion batteries due to their structural tunability and enhanced phase stability. However, a fundamental understanding of how configurational entropy influences the local electronic environment and structural robustness remains elusive. A novel high-configurational-entropy O3-type layered oxide, Na0.85Mn0.40Ni0.25Fe0.1Cu0.1Li0.05Ti0.05Sn0.05O2 (HE-NMNFCLTS), was designed and synthesized as a promising cathode material for sodium-ion batteries. By tailoring the configurational entropy, the crystal field stabilization energy (CFSE) of HE-NMNFCLTS is significantly enhanced: multi-elemental co-doping induces local distortion of MnO6 octahedra, leading to in-creased t2g orbital splitting and a notable improvement in CFSE, thereby stabilizing the crystal structure. Multidimensional characterization reveals its uniform strain distribution, effectively suppressing lattice mismatch and cation migration during Na+ (de) intercalation. Furthermore, HE-NMNFCLTS delivers a high initial discharge capacity of 137.7 mA h g−1 at 0.1 C (2–4.2 V), with 79.5 mA h g−1 retained at 10 C and 80 % capacity retention after 1000 cycles. First-principles calculations further re-veal the enhanced CFSE and a synergistic compensation effect that mitigates anisotropic strain in the TMO2 layers and narrows the bandgap to 0.189 eV, boosting both electronic and ionic conductivity. This work offers a new strategy for stabilizing layered oxide structures via entropy-driven CFSE modulation
Coming to a hard stop? Effect of tapered tocilizumab after weekly tocilizumab cessation for GCA : a multicentre evaluation
No full textObjectives In England, there is a ‘hard stop’ to weekly tocilizumab (qwTCZ) therapy for GCA; this is currently 12 months but was extended during the COVID-19 pandemic subject to certain criteria for GCA relapse risk. Taking advantage of variation in practice, we aimed to compare outcomes of GCA patients who tapered-TCZ vs those who stopped abruptly (non-taper patients). Methods Secondary analysis of an English multicentre service evaluation of relapse after stopping qwTCZ for GCA. Time to relapse was compared between taper and non-taper patients. We examined outcomes according to whether they had been ‘adequate responders’ during qwTCZ therapy, defined as those in remission and on ≤5 mg prednisolone at qwTCZ cessation, without relapse whilst taking qwTCZ. Results We analysed 336 patients from 40 centres. Time to relapse after qwTCZ cessation was significantly longer in adequate responders than non-adequate responders (P = 0.0004). 17.0% (57/336) patients tapered to fortnightly TCZ after qwTCZ cessation, for a median of 6 (IQR 2–13) months. For adequate responders, time to relapse whilst taking tapered-dose TCZ was significantly longer compared with those in the non-taper group (P = 0.0231) based on a relatively small number of flares. There was no difference between the taper and non-tapered groups after tapered-TCZ was stopped (P = 0.8346). In contrast, time to relapse for non-adequate responders was similar in taper patients compared with non-taper patients (P = 0.4892). Conclusion Tapering TCZ after qwTCZ cessation delayed relapse only during the tapering period, but only in adequate responders to qwTCZ. No lasting benefit was seen after tapering ended
Euclid : Early Release Observations – A combined strong and weak lensing solution for Abell 2390 beyond its virial radius
No full textEuclid is currently mapping the distribution of matter in the Universe in detail via the weak lensing signature of billions of distant galaxies. The weak lensing signal is most prominent around galaxy clusters and can extend up to distances well beyond their virial radius, thus constraining their total mass. Near the centre of clusters, where contamination by member galaxies is an issue, the weak lensing data can be complemented with strong lensing data. Strong lensing information can also diminish the uncertainty due to the mass-sheet degeneracy and provide high-resolution information about the distribution of matter in the centre of clusters. Here we present a joint strong and weak lensing analysis of the Euclid Early Release Observations of the cluster Abell 2390 at z = 0.228. Thanks to Euclid ’s wide field of view of 0.5 deg 2 , combined with its angular resolution in the visible band of 0.″13 and sampling of 0.″1 per pixel, we constrained the density profile in a wide range of radii, 30 kpc < r < 2000 kpc, from the inner region near the brightest cluster galaxy to beyond the virial radius of the cluster. We find relatively good consistency with earlier X-ray results based on assumptions of hydrostatic equilibrium and thus indirectly confirm the nearly relaxed state of this cluster. We also find consistency with previous results based on weak lensing data and ground-based observations of this cluster. From the combined strong + weak lensing profile, we derive the values of the viral mass M 200 = (1.48 ± 0.29) × 10 15 M ⊙ , and virial radius r 200 = (2.05 ± 0.13 Mpc), with error bars representing one standard deviation. The profile is well described by a Navarro-Frenk-White model with a concentration c = 6.5 and a small-scale radius of 230 kpc in the 30 kpc < r < 2000 kpc range that is best constrained by strong lensing and weak lensing data. Abell 2390 is the first of many examples where Euclid data will play a crucial role in providing masses for clusters. The large coverage provided by Euclid , combined with the depth of the observations and high angular resolution, will allow us to produce similar results in hundreds of other clusters with rich strong lensing data that is already available
The can challenge : Understanding the best ways to incentivise recycling through a diffusion approach
Get PDFUnderstanding the best ways to incentivise recycling and improve the efficiency of waste practices is a key environmental, social, and economic management problem that needs addressing. We search for solutions to this issue by testing the effectiveness of two incentive mechanisms (a piece-rate and a lottery-based systems). We run a similar field experiment in three different locations, namely a student, residential and workplace environment, to verify the robustness of our findings and thus increase confidence in the external validity of our intervention. By interpreting recycling activity as marketable service, we employ a diffusion model to analyse the potential adoption of the service. We find that monetary rewards increase recycling levels in almost all locations, independently of the type of incentive. More specifically, incentivising recycling stimulates action by those on lower incomes through opportunities for income generation. By contrast, those in workplace environments engage with or without incentives, but the latter provides a boost. Diffusion seems to exist in each environment, but with differing levels of success. Our study contributes to the literature by providing evidence on how to best increase public involvement through recycling and provides important insights for policy making to address this worldwide relevant issue
Functionalized MXene-reinforced PPO-based nanocomposite anion exchange membranes for high-performance water electrolysis
No full textOne of the major challenges in advancing anion exchange membrane water electrolysis (AEMWE) lies in addressing the trade-offs between dimensional stability, electrochemical performance, and mechanical integrity of anion exchange membranes (AEMs). Overcoming these trade-offs is crucial for improving the performance and reliability of AEMWE systems. In this study, we addressed this challenge by developing a series of nanocomposite AEMs by reinforcing quaternized poly(phenylene oxide) (QPPO) with ammonium-functionalized MXene (NH4+-Ti3C2Tx). The resulting membranes exhibited significant enhancement in dimensional stability, electrochemical performance, mechanical strength, and thermal stability compared to the pristine QPPO membrane. Among the composite AEMs, QPPO-NH4-Mx-3.0 demonstrated the notable overall performance, achieving an impressive hydroxide (OH−) ion conductivity (IC) of 153.2 mS cm−1 at 80 °C-2.5 times higher than that of virgin QPPO. Additionally, the membrane contributed excellent tensile strength of 61.2 MPa, 4 times greater than that of the pristine QPPO, and achieved a peak current density of 2.1 A cm−2 at 2 V in 1 M KOH at 60 °C. The membrane also showcased exceptional alkaline stability, retaining 80.2 % of its initial IC after three weeks of immersion in 1 M KOH. Durability testing further validated its robustness, achieved a stable operation maintained up to 150 h of electrolysis with a minimal voltage decay rate of 1.5 mV h−1. In summary, the QPPO-NH4-Mx-3.0 demonstrates substantial promise as a high-performance AEM for advancing AEMWE technology, paving the way for more efficient and reliable water electrolysis systems