The uncertainty of identity toolset:analysing digital traces for user profiling

People manage a spectrum of identities in cyber domains. Profiling individuals and assigning them to distinct groups or classes have potential applications in targeted services, online fraud detection, extensive social sorting, and cyber-security. This paper presents the Uncertainty of Identity Toolset, a framework for the identification and profiling of users from their social media accounts and e-mail addresses. More specifically, in this paper we discuss the design and implementation of two tools of the framework. The Twitter Geographic Profiler tool builds a map of the ethno-cultural communities of a person's friends on Twitter social media service. The E-mail Address Profiler tool identifies the probable identities of individuals from their e-mail addresses and maps their geographical distribution across the UK. To this end, this paper presents a framework for profiling the digital traces of individuals.

Quantitative risk assessment of groundwater quality utilizing GIS technology and coupled groundwater models

The thesis presents a two-dimensional Risk Assessment Method (RAM) where the assessment of risk to the groundwater resources incorporates both the quantification of the probability of the occurrence of contaminant source terms, as well as the assessment of the resultant impacts. The approach emphasizes the need for a greater dependency on the potential pollution sources, rather than the traditional approach where assessment is based mainly on the intrinsic geo-hydrologic parameters. The risk is calculated using Monte Carlo simulation methods whereby random pollution events were generated to the same distribution as historically occurring events or a priori potential probability distribution. Integrated mathematical models then simulate contaminant concentrations at the predefined monitoring points within the aquifer. The spatial and temporal distributions of the concentrations were calculated from repeated realisations, and the number of times when a user defined concentration magnitude was exceeded is quantified as a risk. The method was setup by integrating MODFLOW-2000, MT3DMS and a FORTRAN coded risk model, and automated, using a DOS batch processing file. GIS software was employed in producing the input files and for the presentation of the results. The functionalities of the method, as well as its sensitivities to the model grid sizes, contaminant loading rates, length of stress periods, and the historical frequencies of occurrence of pollution events were evaluated using hypothetical scenarios and a case study. Chloride-related pollution sources were compiled and used as indicative potential contaminant sources for the case study. At any active model cell, if a random generated number is less than the probability of pollution occurrence, then the risk model will generate synthetic contaminant source term as an input into the transport model. The results of the applications of the method are presented in the form of tables, graphs and spatial maps. Varying the model grid sizes indicates no significant effects on the simulated groundwater head. The simulated frequency of daily occurrence of pollution incidents is also independent of the model dimensions. However, the simulated total contaminant mass generated within the aquifer, and the associated volumetric numerical error appear to increase with the increasing grid sizes. Also, the migration of contaminant plume advances faster with the coarse grid sizes as compared to the finer grid sizes. The number of daily contaminant source terms generated and consequently the total mass of contaminant within the aquifer increases in a non linear proportion to the increasing frequency of occurrence of pollution events. The risk of pollution from a number of sources all occurring by chance together was evaluated, and quantitatively presented as risk maps. This capability to combine the risk to a groundwater feature from numerous potential sources of pollution proved to be a great asset to the method, and a large benefit over the contemporary risk and vulnerability methods.

Gram-positive infections in patients with end-stage renal disease

Gram-positive microorganisms, specifically coagulase-negative staphylococci are the most common species recovered from clinical culture specimens of patients with end-stage renal disease. The propensity of coagulase-negative staphylococci (CNS) to cause infection in this patient group has been widely debated. However, it is still unclear how this usually avirulent commensal microorganism produces infection that contributes to high rates of morbidity and mortality in patients with end-stage renal disease. The aim of this thesis was to investigate the rate, geographical distribution, molecular and phenotypic mechanisms of Gram-positive microorganisms associated with infection in renal dialysis patients. In addition, it sought to assess the value of early serological diagnosis of dialysis catheter-associated infection and the effect of antimicrobial treatment regimens on the faecal carriage of enteric microorganisms. In this study, the incidence of haemodialysis catheter-associated infection was established with the Meditrend audit tool. This tool was used to assess the infection outcomes of catheter insertion and management procedures until the catheter was explanted. Introduction of a catheter management protocol decreased the incidence of catheter-related infection. Staphylococcal species recovered from episodes of haemodialysis catheter-associated infection and continuous ambulatory peritoneal dialysis (CAPD)-associated peritonitis were genotyped by determination of macrorestriction profiles with pulsed-field gel electrophoresis. This highlighted horizontal transfer of microorganisms between different patients and the environment. The phenotypic characteristics of these strains were also investigated to determine characteristics that could be used as markers for dialysis catheter-associated infection. The expression of elastase, lipase and esterase by CNS was significantly associated with infection. A rapid enzyme-linked immunosorbent assay incorporating a novel staphylococcal antigen (lipid S) was used to evaluate the early detection of anti-staphylococcal immunoglobulin gamma in patient sera. The comparison of culture positive and culture negative patients demonstrated a steady state of immune activation in both groups. However anti-lipid S serum antibody titres > 1000 were found to be a predictor of infection. The effect on faecal carriage of vancomycin resistant enterococci (VRE) and Clostridium difficile toxins in patients treated with CAPD when empiric cephalosporin therapy was substituted for piperacillin/tazobactam was investigated. The introduction of piperacillin/tazobactam demonstrated a decrease in the faecal carriage of VRE.

Gold mineralisation in the Caledonides of the British Isles with special reference to the Dolgellau Gold-belt, North Wales and the Southern Uplands, Scotland. Available in 2 volumes.

Two aspects of gold mineralisation in the Caledonides of the British Isles have been investigated: gold-telluride mineralisation at Clogau Mine, North Wales; and placer gold mineralisation in the Southern Uplands, Scotland. The primary ore assemblage at Clogau Mine is pyrite, arsenopyrite, cobaltite, pyrrhotine, chalcopyrite, galena, tellurbismuth, tetradymite, altaite, hessite, native gold, wehrlite, hedleyite, native bismuth, bismuthunite and various sulphosalts. The generalised paragenesis is early Fe, Co, Cu, As and S species, and later minerals of Pb, Bi, Ag, Au, Te, Sb. Electron probe micro-analysis (EPMA) of complex telluride-sulphide intergrowths suggests that these intergrowths formed by co-crystallisation/replacement processes and not exsolution. Minor element chemical variation, in the sulphides and tellurides, indicates that antimony and cadmium are preferentially partitioned into telluride minerals. Mineral stability diagrams suggest that during gold deposition log bf aTe2 was between -7.9 and -9.7 and log bf aS2 between -12.4 and -13.8. Co-existing mineral assemblages indicate that the final stages of telluride mineralisation were between c. 250 - 275oC. It is suggested that the high-grade telluride ore shoot was the result of remobilisation of Au, Bi, Ag and Te from low grade mineralisation elsewhere within the vein system, and that gold deposition was brought about by destabilisation of gold chloride complexes by interaction with graphite, sulphides and tellurbismuth. Scanning electron microscopy of planer gold grains from the Southern Uplands, Scotland, indicates that detailed studies on the morphology of placer gold can be used to elucidate the history of gold in the placer environment. In total 18 different morphological characteristics were identified. These were divided on an empirical basis, using the relative degree of mechanical attrition, into proximal and distal characteristics. One morphological characteristic (a porous/spongy surface at high magnification) is considered to be chemical in origin and represent the growth of `new' gold in the placer environment. The geographical distribution of morphological characteristics has been examined and suggests that proximal placer gold is spatially associated with the Loch Doon, Cairsphairn and Fleet granitoids. Quantitative EPMA of the placer gold reveals two compositional populations of placer gold. Examination of the geographical distribution of fineness suggests a loose spatial association between granitoids and low fineness placer gold. Also identified was chemically heterogeneous placer gold. EPMA studies of these heterogeneities allowed estimation of annealing history limits, which suggest that the heterogeneities formed between 150 and 235oC. It is concluded, on the basis of relationships between morphology and composition, that there are two types of placer gold in the Southern Uplands: (i) placer gold which is directly inherited from a hypogene source probably spatially associated with granitoids; and (ii) placer gold that has formed during supergene processes.

Concepts and methods of quantitative assessment of risk to groundwater resources

This work presents a two-dimensional approach of risk assessment method based on the quantification of the probability of the occurrence of contaminant source terms, as well as the assessment of the resultant impacts. The risk is calculated using Monte Carlo simulation methods whereby synthetic contaminant source terms were generated to the same distribution as historically occurring pollution events or a priori potential probability distribution. The spatial and temporal distributions of the generated contaminant concentrations at pre-defined monitoring points within the aquifer were then simulated from repeated realisations using integrated mathematical models. The number of times when user defined ranges of concentration magnitudes were exceeded is quantified as risk. The utilities of the method were demonstrated using hypothetical scenarios, and the risk of pollution from a number of sources all occurring by chance together was evaluated. The results are presented in the form of charts and spatial maps. The generated risk maps show the risk of pollution at each observation borehole, as well as the trends within the study area. This capability to generate synthetic pollution events from numerous potential sources of pollution based on historical frequency of their occurrence proved to be a great asset to the method, and a large benefit over the contemporary methods.