OAK (Online Access to Knowledge) Commons (Young Harris College)
Not a member yet
33676 research outputs found
Sort by
Natural Language Inference over Tables: Enabling Explainable Data Exploration on Data Lakes
No full textData lakes are repositories of data with potential for analysis. Data lakes aim to liberate data from silos, thereby enabling cross-cutting analyses that were hitherto out of reach. This gives rise to significant challenges for data scientists simply discovering what data sets may be relevant to a task-in-hand. Given a data set of interest, several proposals have been made for indexing schemes that can identify related data sets. However, such schemes tend to build on similarity metrics that stop short of providing a clear explanation as to how an identified data set relates to a provided target. We address this problem by applying Natural Language Inference (NLI) to providing explanations as to how the attributes of discovered data sets relate to those of the target, in terms of a collection of semantic relations. We provide two approaches to inferring semantic relations: (a) by performing unsupervised intensional and extensional analysis of the data sources using Natural Language Processing techniques; and (b) by performing supervised learning of semantic relations by applying BERT over source schema information. The contributions of this paper are: an NLI strategy for providing explicit characterisation of semantic relations between data sets; two approaches to inferring the semantic relations; and an empirical evaluation of the approaches using open government data
Generating Three-dimensional Drape Model Based on Projection and Mesh Deformation
No full textIn order to generate the three-dimensional (3D) mesh of draped fabric based on a two-dimensional (2D) fabric projection, the 3D point cloud of draped fabric was scanned via a self-built 3D scanning device followed by triangulation. The 3D boundary of fabric drape model and the two-dimensional contour of the projection of the fabric drape model were extracted respectively. Two neural networks were constructed to bridge the 2D fabric drape projection and the 3D boundary of draped fabric. With the trained neural networks, the 3D boundary of draped fabric could be inferred by merely referring a 2D projection of draped fabric. Six reference triangular meshes with different triangle density were generated. Driven by two mesh deformation methods, Laplacian deformation and Poisson deformation, the reference mesh was deformed into 3D mesh similar to scanned fabric drape models (meshes). The result shows that the generated models have a high agreement with the scanned meshes. In addition, the Laplacian deformation outperforms Poisson deformation in generating 3D fabric drape models
Recent advances in phosphoric acid–based membranes for high–temperature proton exchange membrane fuel cells
No full textHigh‐temperature proton exchange membrane fuel cells (HT‐PEMFCs) are pursued worldwide as efficient energy conversion devices. Great efforts have been made in the area of designing and developing phosphoric acid (PA)–based proton exchange membrane (PEM) of HT‐PEMFCs. This review focuses on recent advances in the limitations of acid–based PEM (acid leaching, oxidative degradation, and mechanical degradation) and the approaches mitigating the membrane degradation. Preparing multilayer or polymers with continuous network, adding hygroscopic inorganic materials, and introducing PA doping sites or covalent interactions with PA can effectively reduce acid leaching. Membrane oxidative degradation can be alleviated by synthesizing crosslinked or branched polymers, and introducing antioxidative groups or highly oxidative stable materials. Crosslinking to get a compact structure, blending with stable polymers and inorganic materials, preparing polymer with high molecular weight, and fabricating the polymer with PA doping sites away from backbones, are recommended to improve the membrane mechanical strength. Also, by comparing the running hours and decay rate, three current approaches, 1. crosslinking via thermally curing or polymeric crosslinker, 2. incorporating hygroscopic inorganic materials, 3. increasing membrane layers or introducing strong basic groups and electron–withdrawing groups, have been concluded to be promising approaches to improve the durability of HT–PEMFCs. The overall aim of this review is to explore the existing degradation challenges and opportunities to serve as a solid basis for the deployment in the fuel cell market
Energy efficient production of 5-hydroxymethylfurfural (5-HMF) over surface functionalized carbon superstructures under microwave irradiation
No full textMicrowave (MW)-assisted process intensification of catalytic conversion of saccharides is a promising route for energy efficient production of value-added chemicals and fuels. This work presents the development of the MW-responsive catalysts (i.e., the acid functionalized spherical carbon superstructure, SCS) for converting fructose to 5-HMF effectively and efficiently under MW irradiation. Under a mild MW condition (at 70 C, atmospheric pressure), the SCS catalyst (with surface sulfonic and carboxylic acid groups) achieved fructose conversion of ~97.8% and selectivity to 5-HMF of ~90.6% after 5 min reaction time with good recyclability, being significantly higher than that obtained by the system under the conventional heating. Importantly, the MW-assisted system showed very high energy efficiency coefficient of 1.01 mmol kJ−1 L−1 compared to that (0.09 mmol kJ−1 L−1) of the conventional thermal catalysis. The mechanism of process intensification of the system developed may be attributed to the combination of the dipole polarization of the MW-responsive sulfonic groups and the selective heating of the MW-absorbing SCS support under the MW condition
Duality bounds for discrete-time Zames–Falb multipliers
No full textThis note presents phase conditions under which there is no suitable Zames–Falb multiplier for a given discrete time system. Our conditions can be seen as the discrete-time counterpart of Jönsson’s duality conditions for Zames–Falb multipliers. By contrast with their continuous-time counterparts and other phase limitations in the literature, they lead to numerically efficient results that can be computed either in closed form or via a linear program. The closed-form phase limitations are tight in the sense that we can construct multipliers that meet them with equality. The numerical results allow us to conclude that the current state-of-the-art in searches for Zames–Falb multipliers are not conservative. Moreover, they allow us to show, by construction, that the set of plants for which a suitable Zames– Falb multiplier exists is non-convex
Structuralist Analysis For Neural Network System Diagrams
No full textThis paper examines diagrams describing neural network systems in academic conference proceedings. Many aspects of scholarly communication are controlled, particularly with relation to text and for- matting, but often diagrams are not centrally curated beyond a peer review. Using a corpus-based approach, we argue that the heterogeneous diagrammatic notations used for neural network systems have implications for signification in this domain. We divide this into two questions (i) what content is being represented and (ii) how relations are encoded. Using a novel structuralist framework, we use a corpus analysis to quantitatively cluster diagrams according to the author's representational choices. This quantitative diagram classification in a heterogeneous domain may provide a foundation for categorising representational properties of diagrams
Microrheology of colloidal suspensions via Dynamic Monte Carlo simulations
No full textUnderstanding the rheology of colloidal suspensions is crucial in the formulation of a wide selection of industry-relevant products, such as paints, foods and inks. To characterise the viscoelastic behaviour of these soft materials, one can analyse the microscopic dynamics of colloidal tracers diffusing through the host fluid and generating local deformations and stresses. This technique, referred to as microrheology, links the bulk rheology of fluids to the microscopic dynamics at the particle scale. If tracers are subjected to external forces, rather than freely diffusing, it is called active microrheology. Motivated by the impact of microrheology in providing information on local structure in complex systems such as colloidal glasses, active matter or biological systems, we have extended the dynamic Monte Carlo (DMC) technique to investigate active microrheology in colloidal suspensions. The original DMC theoretical framework, able to accurately describe the Brownian dynamics of colloids at equilibrium, is here reconsidered and expanded to describe the effects of an external force pulling a tracer embedded in isotropic colloidal suspensions at different densities. To this end, we studied the dynamics of a spherical tracer dragged by a constant external force through a bath of spherical and rod-like particles of comparable size. We could extract valuable details on its effective friction coefficient, being constant at small and large values of the external force, but otherwise displaying a nonlinear behaviour that indicates the occurrence of a force-thinning regime. Our DMC simulation results are in excellent quantitative agreement with past Langevin dynamics simulations and theoretical works for the bath of spherical colloids. The bath of rod-like particles is studied in the isotropic phase, and displays an example where DMC is more convenient than Brownian or Langevin dynamics, in this case in dealing with particle rotation
Determinants of eclampsia in women with severe preeclampsia at Mpilo Central Hospital, Bulawayo, Zimbabwe.
No full textObjective Globally, preeclampsia is a significant contributor to adverse maternal outcomes. Once women develop eclampsia, they face considerable risks especially in countries with limited resources to deal with such a life-threatening complication. This study was carried out to investigate determinants of eclampsia in pregnant mothers with severe preeclampsia. Study design This institutional based study was completed at Mpilo Central Hospital, a quaternary referral unit from 1st January 2016 – 31st December 2018. In this study, pregnant women with severe preeclampsia/eclampsia were the study participants. The independent variables included socio-demographic and clinical characteristics, and maternal outcomes. Multivariable logistic regression analyses were used to determine independent association with p < 0.05 taken as statistically significant with 95% Confidence Interval (CI). Main outcome measure Eclampsia Results Development of eclampsia was more frequent in women aged 14-19 years compared to women aged ≥35 years (adjusted odds ratio (AOR) 6.64, 95% CI 1.20-22.06, p=0.02) and in primiparous women compared to women with parity ≥3 (AOR 2.76, 95% CI 1.48-5.15, p=0.001). Eclampsia was more frequent in women with diastolic blood pressure of 131-150mmHg (AOR 5.48, 95% CI 1.05-28.75, p=0.04), and ≥150mmHg (AOR 5.78, 95% CI 1.05-31.78, p=0.04) compared with those with diastolic blood pressure of ≤110mmHg. Symptoms of visual disturbances were also associated with eclampsia (AOR 2.13, 95% CI 1.08-4.18, p=0.03). Conclusions This study has identified independent determinants of eclampsia which can be used to identify which women should receive magnesium sulphate prophlyaxis or more intensive monitoring to prevent deterioration in maternal condition
Controlled trials and meta-analyses of electroconvulsive therapy (ECT) against sham ECT for depression: do study limitations invalidate the evidence (and mean we should stop using ECT)?:Commentary on … Electroconvulsive Therapy for Depression: A Review of the Quality of ECT versus Sham ECT Trials and Meta-Analyses, Read et al (2020)
No full textElectroconvulsive therapy (ECT) for depression is a controversial treatment with highly polarised views about the balance between therapeutic benefits and adverse effects. Studies investigating whether ECT is more effective than a placebo treatment started in the 1950s, with the most important randomised controlled studies carried out about four decades ago in which ECT was compared with sham ECT involving anaesthesia but no electrically-induced seizure. Subsequently the data have been pooled in a number of meta-analyses which have found that ECT is an effective treatment. However a recent review of the quality of the sham ECT-controlled studies, and the meta-analyses based on them, concludes that their quality is too poor to allow assessment of the efficacy of ECT, and that given its risks (permanent memory loss and death) the use of ECT should be suspended. This commentary critically discusses the methodology of this review and its conclusions
Sorption, swelling and plasticization of PIM-1 in methanol-dimethyl carbonate vapour mixtures
No full textMembranes from PIM-1 represent a promising tool for the efficient recovery of organic components from their vapour mixtures with water or alcohols. To understand the phenomena governing the separation, sorption, swelling and mechanical characteristics were systematically studied for thick PIM-1 films exposed to vapour mixtures of methanol and dimethyl carbonate (DMC). The studied binary vapour mixtures contained 30, 55, 82 and 90 mol.% of methanol and reached 28, 49 and 63 % of the dew point pressure at 40 °C. The separation factor for sorption of vapour mixtures ranged 4.0-7.5, peaked for the azeotropic mixture (82 mol% of methanol) at low saturation (28 % of the dew point pressure), and was significantly contributed by competitive sorption. Furthermore, the tested PIM-1 films exhibited anomalously low volume swelling while their storage modulus remained comparable to that of the pure polymer until exposed to the highly saturated vapours (63 % of the dew point pressure). Raman spectroscopy analysis of PIM-1 swollen by liquid methanol and DMC revealed the pronounced frequency shifts of the C-C vibrations in the aromatic rings and the C-H vibrations in the pentacyclic units. These centres presumably triggered the relaxation of the polymer backbone occurring at high saturations of the vapours. In conclusion, PIM-1 remains highly rigid and selective to DMC in wide ranges of saturation and composition for vapour mixtures of methanol with DMC and is a candidate for effective inversely selective vapour separation membranes