Among tourism decision making studies, choice heuristics are barely explored. Choice heuristics are a set of rules decision makers use to process information about their choice criteria. The types of rules applied in the process can make a substantial difference to eventual outcomes, which for tourists includes not only the final choice of the tour package but also the choice process (Sen, 1997). Therefore, in order to better understand tourists' decision making behaviour, choice heuristics deserve greater attention. However, whilst compensatory choice heuristics are commonly applied in consumer focused tourism research, and conjoint analysis is often used for modelling compensatory heuristics, mirroring the predominant approach in consumer research, less attention has been afforded to non-compensatory heuristics. Recently, a greedoid method was introduced by Kohli and Jedidi (2007) and Yee et al. (2007) independently to that contributed to methodology for estimating non-compensatory heuristics. The aim of this study was to provide a greater understanding of consumer decision-making processes, based on the exploration of different choice heuristics used by the Chinese long-haul outbound tourists. This thesis makes theoretical contribution by providing insights on (1) how the concept of choice heuristics can be used to better understand the process of decision making and (2) how choice heuristics are used for the selection of complicated intangible services, tourism destinations in this case. The study also sheds light on the possible measurements for evaluating the fit of different choice heuristic models. In addition, the information found regarding the destination preference of Chinese long-haul outbound tourist is of great use for marketers and suppliers to improve their destination products as well as their advertisement campaigns.

A software system has been developed that monitors and interprets temporally changing (internal) building environments and generates related knowledge that can assist in facility management (FM) decision making. The use of the multi agent paradigm renders a system that delivers demonstrable rationality and is robust within the dynamic environment that it operates. Agent behaviour directed at working toward goals is rendered intelligent with semantic web technologies. The capture of semantics though formal expression to model the environment, adds a richness that the agents exploit to intelligently determine behaviours to satisfy goals that are flexible and adaptable. The agent goals are to generate knowledge about building space usage as well as environmental conditions by elaborating and combining near real time sensor data and information from conventional building models. Additionally further inferences are facilitated including those about wasted resources such as unnecessary lighting and heating for example. In contrast, current FM tools, lacking automatic synchronisation with the domain and rich semantic modelling, are limited to the simpler querying of manually maintained models.

Rotation is a key physical process operating in the A stars. We present a method by which rotation periods might be measured with Fourier transforms of Kepler data, potentially removing the requirement of spectroscopic observations to acquire similar information. Angular rotational velocities reach their maximum at about A5, but slowly rotating A stars are also seen and these tend to be chemically peculiar. For Ap stars, rotational braking is understood, but too few progenitors are observed. A review of a rare class of stars, the ‘sn’ stars, leads us to suggest these may contribute to the ‘missing’ Ap progenitors. For Am stars, we find the tidal braking mechanisms proposed in the literature induce mixing that is incompatible with observed abundance anomalies. At the other end of the scale are the λ Boo stars, whose rotation velocities are above average. The two main theories for the origin of their metal underabundances are discussed and it is suggested that both of them imply the λ Boo stars contain a high fraction of pulsators – a suggestion that is backed up by observations in the literature. Many λ Boo stars also have circumstellar material, suggesting they are potential planet hosts. This, and the use of asteroseismology to study their interiors, are two excellent reasons to adopt them as prime targets for detailed investigation with Kepler. Pulsation is a common phenomenon in A stars. The δ Sct stars receive wide attention but the fraction of stars that pulsate at the 50 μmag level is shown to be only 56 per cent. The non-δ Sct stars in the δ Sct instability strip receive far less attention. Some of these stars, without appreciable granulation or stellar winds, are probably the least variable objects on the HR diagram. It is shown that they have the potential to be among the most peculiar Am stars. Their investigation has led to two important conclusions: (1) the presence of γ Dor pulsation in non-δ Sct stars in the δ Sct instability strip may inhibit the development of the expected Am peculiarities; and (2) chemically normal, non-δ Sct stars in the δ Sct instability strip do exist at the μmag level. Fourier transforms are invaluable tools in the field of stellar pulsation. The Kepler Space Telescope is providing data of exquisite precision, and thus more detail is seen in the Fourier transform than ever before. Truly understanding the properties of the data is fundamental to their successful utilisation. Through statistical analysis of noise levels in over 20 000 stars, granulation is concluded to be visible in Fourier transforms of stars cooler than 7500 K. Another property investigated is the Nyquist frequency. We found that periodically modulated sampling on board Kepler allows distinction between real pulsation frequencies and Nyquist aliases, even when those real frequencies exceed the Nyquist frequency of the data. This discovery opens up study of many hundreds of stars previously thought to have insurmountable Nyquist ambiguities in their data.

This research has investigated marine biofouling and its influence on the durability of concrete sea defences using on-site and laboratory-based studies. The study was divided into three main phases namely: the surface analysis of armour concrete, the study of algal colonisation within the matrix and investigations into the presence of a bacterial biofilm within freshly hardened armour concrete. The effectiveness of photocatalytic coatings as a non-toxic anti-fouling strategy and cell attachment to synthetic fibres was also studied. It was found that algal growth quickly developed at the interface of inclusions within the matrix and that power washing with the use of Dairy Hypochlorite to remove this accelerated wear, leading to significant mass loss. It was also observed that bacterial growth within local beach sand, which was used in the production of the revetment armour units, survived the concrete manufacturing process. Bacteria were cultured from the sand and were found to match the Actinomycete like growth in the freshly hardened matrix of armour concrete. This thesis proposes a holistic model for biofouling of fibre reinforced marine concrete in which algal growth around inclusions facilitates a complex process of biodeterioration. Bacterial filamentous growth around and through synthetic fibres embedded in the new concrete mix, appears to be detrimental to the long term durability of synthetic fibres. Subsequent algal colonisation on the surface of newly placed units appeared to quickly penetrate the surface through exposed fibres and percolated interfaces of inclusions, subsequently weakening their bond. During the manufacture of the armour units, aggregate segregation in the 90° corners in the bottom of the form created a weaker matrix in the surface region most exposed to biodeterioration, the full force of wave action and power washing. The main conclusions from this study are: • Synthetic fibres used at the study site are inappropriate for marine concrete, particularly in algal rich waters, within the inter-tidal zone where beach sand is used in the concrete mix. Amendments to Concrete Society Technical Report No. 65: Guidance on the use of Macro-synthetic-fibre-reinforced concrete have been recommended. • Bacterial loaded beach sand is detrimental to the durability of marine concrete in the inter-tidal zone and amendments are recommended to PD 6682-1:2013 Aggregates for concrete (BSI, 2013a) in order to highlight this concern. This UK guidance suggests limiting values for aggregate properties within the ranges permitted in BS EN 12620 (BSI, 2013) but does not place any limits on microorganisms present in beach sand. Further work is needed into the susceptibility of synthetic fibres to crystal growth. Alterations in the manufacture of armour units have been recommended by this author.

Structured photonic materials are typically composed of periodic subwavelength elements where the unit cell geometries can impact the overall optical characteristics of the bulk material. By using micro and nanofabrication technologies it is possible to engineer the electromagnetic properties of structured photonic materials for a given application and create a variety of optical components such as band pass filters and absorbers. Two structured photonic materials that have gained substantial interest in recent years are plasmonic filters and metamaterials which are well suited for optical and terahertz imaging applications, respectively. In addition to imaging applications within individual wavebands, structured photonic materials, such as plasmonic filters and metamaterials, could be hybridised and combined with suitable sensors to create a multi-spectral imaging system capable of imaging at optical and terahertz wavebands simultaneously. These new hybrid structured photonic materials are known as synthetic multi-spectral materials, and their development will be presented in this work. To design synthetic multi-spectral materials it was necessary to optimise the plasmonic filter and metamaterial components independently. This involved electromagnetic simulation studies using finite-difference time-domain techniques, fabrication of the structured materials and characterisation using suitable techniques for the relevant spectral band. It was also necessary to ensure that all structures used the same materials and similar fabrication processing techniques as a means of simplifying hybridisation of the two structures. Plasmonic filters exhibit extraordinary optical transmission due to coupling of light with surface plasmons at a metal-dielectric interface. A 16 colour plasmonic filter set, consisting of triangular hole arrays etched into an aluminium film, was optimised for imaging applications in the visible and near infrared spectral range. Initial work on the integration of synthetic multi-spectral materials with CMOS image sensors was undertaken by developing fabrication processes to integrate plasmonic colour filters with two different CMOS chips. Preliminary results from the characterisation of the optical filters fabricated on to the chips have been presented. The resonant wavelengths of the plasmonic colour filters were then scaled up to infrared wavelengths where it was necessary to consider the role of spoof surface plasmons on the extraordinary optical transmission phenomenon. This led to the fabrication of 8 short wave infrared plasmonic filters. Metamaterial band pass filters consist of a single metal film etched with a periodic complementary electric ring resonator unit cell structure. Metamaterial absorbers consist of an electric ring resonator, separated by a metallic ground layer by a dielectric spacer. In the course of this work, two metamaterial filters and four metamaterial absorbers were designed. The metamaterial structures exhibit resonant characteristics at terahertz frequencies. Three synthetic multi-spectral materials, each consisting of hybrid plasmonic filter and terahertz metamaterial structures, have been simulated, fabricated and characterised. The first synthetic multi-spectral material combines 16 plasmonic filters with a terahertz metamaterial filter and is capable of filtering 15 optical wavelengths and a single near infrared wavelength, whilst simultaneously filtering a single terahertz frequency. The multi-spectral filter demonstrates that it is possible to engineer the optical passband characteristics of a thin metal film over several decades of wavelength using a single electron beam lithography step. The second synthetic multi-spectral material consists of 16 plasmonic filters hybridised with a terahertz metamaterial absorber and can filter 15 optical wavelengths and a single near infrared wavelength whilst simultaneously absorbing a single terahertz frequency. Plasmonic filters and metamaterial absorbers are promising components for use in the development of new optical and terahertz imaging systems, respectively, and therefore the second synthetic multi-spectral material represents a significant step forward in the development of a visible and terahertz multi-spectral camera. The third synthetic multi-spectral material combines 7 plasmonic filters with a low metal fill factor metamaterial absorber, to increase the measured transmission of the plasmonic filter components. The third synthetic multi-spectral material is capable of filtering three optical wavelengths, a single near infrared wavelength, a single short wave infrared wavelength and two mid infrared wavelengths, whilst simultaneously absorbing a single terahertz frequency. Such a synthetic multi-spectral material could aid in the development of a visible, infrared and terahertz multi-spectral camera.

Technology transfer is extensively believed to be one of the major debates in the literature on development economics. The experiences of some successful countries in rapid economic and industrial development, in particular, some East-Asian Newly Industrialised Countries (NICs) show that the acquisition of a significant amount of foreign technology has played a crucial role. This crucial role includes promoting their managerial and technical expertise as well as increasing their productivity level through the adoption of a set of appropriate policies and strategiesT. hesee xperiencesc ould have valuable lessonsf or other countriesw ho wish to follow similar strategies to achieve rapid industrialisation and technological development. Although many Less Developed Countries (LDCs) have realised the great importance of technological transformation for their rapid economic and industrial development, they have not designed effective and efficient policies for the transfer of appropriate and high-level technologies. The present empirical investigation is intended to contribute to the large existing literature on technological transfer and the role that Multinational Corporations (MNCs) play in this. Its major contribution lies in demonstrating rigorously that the integration of foreign technologies is greatly affected by the socio-economic conditions of the recipient countries. The present study attempts to identify the main socio-economic characteristics of countries involved in assimilating transferred technology. It first identifies the critical success or failure factors for effective technology transfer and the rapid industrialisation of the LDCs in general. Then, it provides a quantifiable metric index of the rate of the technological absorption. Selectiono f relevant variablesa nd choosingt he sampleo f countries are summarisedT. he model, which is based on the multiple regression analysis as well as other statistical techniques, is identified. The four-variable-model derived from the stepwise regression results gave a statistically significant R-sq = 70.71% and R-sq (adj) = 66.7% and satisfies the principle of parsimony, was chosen as the preferred model. This has as explanatory variables transport and communications and gross national savings as economic indicators - Christian religion and natural disasters (negative concept) as social indicators. The results suggest that countries with the above indicators are more able to absorb and integrate foreign technologies. In general, the results reveal that the rate of technology integration varies greatly with the level of socio-economic development. Some intangible factors that cannot as yet be quantified and may be expected to have significant effects on the rate of technological integration, such as political and managerial factors are discussed. The analysis of results is concludedw ith somer ecommendationsa nd suggestionsd erived from the research findings and results for the effective and successful technology transfer of LDCs along with the technology transfer in Africa, problems of AIDS and its impact on African development.

Doped zinc oxide is of interest as a transparent conductive oxide (TCO), due to the abundance of its major constituents, its low resistivity, high transparency and wide bandgap. The current work focuses on the properties required for TCO applications including resistivity of ≤10-3 Ω•cm, carrier density of ≥1020 cm-3, and transparency >80% in the visible light. Zirconium (Zr4+) was chosen as the dopant in the current work due to its abundance, comparable ionic size to Zn and because it can act as a double donor providing up to two extra free electrons per ion when substituted for Zn2+. The doping process can be controlled using atomic layer deposition (ALD), with the doped films resulting in an increased conductivity. The films in the current work resulted in a minimum resistivity of 1.44×10-3 Ω•cm and maximum carrier density of 3.81×1020 cm-3 for films <100 nm thickness, having 4.8 at.% Zr concentration. The resistivity was further reduced after reducing the interfacial and grain boundary scattering (i.e. increase grain size), by increasing the overall film thickness. The resistivity of 7.5×10-4 Ω•cm, carrier mobility of 19.6 cm2V‒1s-1 and carrier density of 4.2×1020 cm-3 were measured for a 250 nm thick film with 4.8 at.% doping. The tuning of the carrier density via doping offers control over the optical gap due to the net effect of Burstein-Moss effect and bandgap renormalisation. This resulted to an increase of the optical gap from 3.2 eV for the un-doped ZnO to 3.5 eV for 4.8 at.% Zr-doped films. The average optical transparency in the visible/near IR range was as high as 91% for 4.8 at.% doped films. The thickness increase also resulted in a grain orientation shift from perpendicular to the substrate (i.e. polar c-plane orientation) to parallel (i.e. non-polar m-plane) due to the strain increase that forced the films to grow at a low strain energy direction. This offers the possibility of growing non-polarised films that show no piezoelectric field charge observed in polar oriented films. Therefore, controlling the grain size through the number of ALD cycles can effectively result in mobility and preferred orientation control, while the doping concentration controls the resistivity, optical bandgap and transparency of the films.

Protein interactions are a fundamental part of cellular processes, and represent a key target in the understanding of cell behaviour, communication and function. These interactions are dynamic in nature, and change over time. Furthermore, interactions can be dependent on co-localisation in cellular compartments. Both of these characteristics can be obscured through observation by bulk molecular cell assays such as co-immunoprecipitation (co-IP) and can obscure the intricacies of single cell dynamics. The development of tools and new methodologies to observe single cell protein interaction dynamics are key to understanding the underlying mechanisms that direct cell fate. Systems biology aims to incorporate into predictive models of the whole system. In this way, the development of quantitative experimental tools is a key component of systems biology. Förster Resonance Energy Transfer is a widely used technique in the field of protein-protein interaction studies. Dependent on the non-radiative transfer of energy from a fluorescent molecule of higher excitation energy to a fluorescent molecule of lower excitation energy with sufficiently overlapping spectra, the process occurs across a 1-10nm (100Å) range. As a result, FRET interactions between fluorophores attached to biologically functional proteins are a strong indication of protein interaction. The photoswitchable protein Dronpa offers a unique opportunity to develop a real-time live cell variant of this technique. Through the modulation of Dronpa fluorescence, repeated donor quenching can be observed, allowing quantification of FRET interactions between fluorophores without spillover. This is achieved sequentially through the optimisation of Dronpa imaging parameters for live cell imaging, followed by the identification and testing of candidate FRET partners using an optimised imaging protocol. Direct fusions were used to qualify potential FRET responses between tested pairs of fluorophores. FRET responses using positive control constructs were rigorously tested and quantified to ensure repeatable and consistent reporting of FRET. The spectral properties of Dronpa and chosen FRET partners were also rigorously tested to ensure FRET could be accurately measured through sensitised emission. Following the confirmation of a reliable FRET response using positive control constructs, the assay was applied to the NF-κB protein p105, and other interacting family members to measure any changes in protein dynamics. The system was applied both on a single switch basis to make comparisons between different combinations of co-expressed fusion proteins and, in time series experiments, to measure potential changes in dynamics over time. p105 showed interesting behaviour not reported in the literature, specifically the detection of full length p105 in the nucleus before stimulation and the strong intramolecular interaction of N and C terminal regions which become perturbed by co-expression of p65. p65 displayed a stronger interaction with the c-terminal region of p105, indicating preferential binding of p65 with ankyrin repeat region rather than the Rel homology domain of p105. Fluorescence Cross Correlation Spectroscopy (FCCS) was used as a complimentary technique to confirm findings in the FRET assay. Considered together, these data suggest Dronpa is a viable component for the accurate detection of FRET in real time, free from complication by spillover. The use of this technique could help to elucidate dynamic protein interactions key to the development and improvement of existing models in systems biology approaches.

Nonnegative matrix factorization (NMF) has becoming an increasingly popular data processing tool these years, widely used by various communities including computer vision, text mining and bioinformatics. It is able to approximate each data sample in a data collection by a linear combination of a set of nonnegative basis vectors weighted by nonnegative weights. This often enables meaningful interpretation of the data, motivates useful insights and facilitates tasks such as data compression, clustering and classification. These subsequently lead to various active roles of NMF in data analysis, e.g., dimensionality reduction tool [11, 75], clustering tool[94, 82, 13, 39], feature engine [40], source separation tool [38], etc. Different methods based on NMF are proposed in this thesis: The modification of k- means clustering is chosen as one of the initialisation methods for NMF. Experimental results demonstrate the excellence of this method with improved compression performance. Independent principal component analysis (IPCA) which combines the advantage of both principal component analysis (PCA) and independent component analysis (ICA) has been chosen as the significant initialisation method for NMF with improved clustering accuracy. We have proposed the new evolutionary optimization strategy for NMF driven by three proposed update schemes in the solution space, saying NMF rule (or original movement), firefly rule (or beta movement) and survival of the fittest rule (or best movement). This proposed update strategy facilitates both the clustering and compression problems by using the different system objective functions that make use of the clustering and compression quality measurements. A hybrid initialisation approach is used by including the state-of-the-art NMF initialization methods as seed knowledge to increase the rate of convergence. There is no limitation for the number and the type of the initialization methods used for the proposed optimisation approach. Numerous computer experiments using the benchmark datasets verify the theoretical results, make comparisons among the techniques in measures of clustering/compression accuracy. Experimental results demonstrate the excellence of these methods with im- proved clustering/compression performance. In the application of EEG dataset, we employed several standard algorithms to provide clustering on preprocessed EEG data. We also explored ensemble clustering to obtain some tight clusters. We can make some statements based on the results we have got: firstly, normalization is necessary for this EEG brain dataset to obtain reasonable clustering; secondly, k-means, k-medoids and HC-Ward provide relatively better clustering results; thirdly, ensemble clustering enables us to tune the tightness of the clusters so that the research can be focused.

The large availability of Kraft lignin as an industrial by-product and its polyaromatic characteristic, is ideal to consider the potential for recycling it into fine chemicals. To depolymerise lignin, solvolysis and hydrogenolysis experiments were performed. This research considered whether the low yields of products (fine chemicals) were related to the low content of β-O-4 bonds or if it was also associated to the dissolution of lignin in the solvent solution employed in the reactions. The type of solvents chosen to check the dissolution effect were those with low cost and were more sustainable than traditional solvents. Water, ethanol, isopropanol (IPA) and acetone were used. The water mixtures were applied in the tests in various proportions (25:75, 50:50, 75:25 solvent/water v:v). Due to their ability to break C-C and C-O bonds in lignin model compounds [1][2], the efficiency of platinum and rhodium in these reactions supported on alumina was also studied. It was found that the non-catalysed (solvolysis) and catalysed reactions showed different selectivities but similar overall yields ~ 10 % wt of monomeric phenols. The difficulty in increasing yields was mainly associated with the highly condensed character of Kraft lignin and re-polymerisation issues. To achieve an understanding of Kraft lignin depolymerisation, isotopic labelling reactions were completed in the presence of deuterated solvents as well as deuterium gas. This gave information on how Kraft lignin depolymerises, the influence of solvent to products formation and the involvement of hydrogen in the rate determining steps in the reactions. These results have led to an initial mechanistic understanding on how this complex molecule may yield alky-phenolic compounds. It was revealed that the solvent was directly involved in the products’ formation and that they were not generated by simple thermolysis. In addition, the presence of catalysts and hydrogen influenced product formation. The compounds showed different kinetic isotopic values, suggesting that each of these molecules came from individual mechanisms, highlighting the complexity of their formation. This was a relevant study as most of lignin depolymerisation mechanistic insights are based on model compounds and not on lignin itself. It was of interest to this project to explore not only different catalysts and their relationship to lignin depolymerisation, but also different lignin types. A simple pre-treatment for lignin extraction using sawdust (from oak and birch wood) in a Parr autoclave reactor in the presence of hydrogen, solvent and high temperature was developed. The lignins obtained after the pre-treatment were named parr-lignin and successfully resulted in polyaromatic molecules with less condensed character compared to lignins from Soda or Kraft pulping. Reactions were carried out with these lignins and a sugar-cane lignin. 4 5 Different catalytic systems with these lignins were investigated and how depolymerisation was affected by the metal and support used. The catalysts involved in the reactions included platinum, rhodium, nickel and iron. Various supports such as alumina, zirconia and carbon were tested along with the metals described. It was found that the supports were not inert in these experiments presenting catalytic activity. Materials with low surface area (zirconium catalysts) gave a poor performance compared to the others. In addition, nickel, a non-noble metal, showed as good a catalytic effect in the depolymerisation of these lignins as Pt and Rh. The components in the system influenced the reactions to different extents, especially product distribution. The catalysts had different selectivities and the solvents were not only dissolving lignin but also influencing the results. GPC analysis was performed to give an overview of the condensed level of these lignins and degrees of depolymerisation compared to the original material. GC-MS enabled the identification and quantification of 18 monomeric compounds. The post reaction characterisation of selected alumina catalysts (Pt/Al2O3, Ni/Al2O3 and Al2O3) was performed using XRD, BET, CHN, TPO and Raman Analysis to study the nature of the carbonaceous layer deposited on these materials. The work showed that after reaction the catalysts turned black in colour and the carbon laydown consisted of not only one simple type of carbon, and included graphitic species. The amount of carbon deposited depended on the type of lignin. Oak and birch parr-lignins had the highest and lowest amount of carbon over the catalysts respectively. No obvious trend relating to the type of catalyst, lignin and solvent used to the carbon nature was identified. This work showed that lignins with less condensed nature were less susceptible to solvolysis and more to hydrogenolysis. For example, sugar-cane lignin gave 3.9% of phenolic compounds in the solvolysis while reaction with Rh/Al2O3 gave 12.9% of products. This indicated that more selective cleavage of bonds were promoted by heterogenous catalysts. The results suggested that some compounds were mainly generated via dealkylation and hydrodeoxygenation, allowing a future possibility to generate target molecules. These results were mainly due to the presence of more labile bonds, vulnerable to hydrogenolysis. Highlighting that prior to depolymerisation, the pre-treatment used to extract lignin must be appropriate to avoid depletion of the alkyl-aryl ether bonds (β-O-4 bonds, especially) relevant for fine chemicals generation.