A two-stage item recommendation method using probabilistic ranking with reconstructed tensor model

In a tag-based recommender system, the multi-dimensional correlation should be modeled effectively for finding quality recommendations. Recently, few researchers have used tensor models in recommendation to represent and analyze latent relationships inherent in multi-dimensions data. A common approach is to build the tensor model, decompose it and, then, directly use the reconstructed tensor to generate the recommendation based on the maximum values of tensor elements. In order to improve the accuracy and scalability, we propose an implementation of the -mode block-striped (matrix) product for scalable tensor reconstruction and probabilistically ranking the candidate items generated from the reconstructed tensor. With testing on real-world datasets, we demonstrate that the proposed method outperforms the benchmarking methods in terms of recommendation accuracy and scalability.

The affect heuristic and public support for three types of wood smoke mitigation policies

This study applied the affect heuristic model to investigate key psychological factors (affective associations, perceived benefits, and costs of wood heating) contributing to public support for three distinct types of wood smoke mitigation policies: education, incentives, and regulation. The sample comprised 265 residents of Armidale, an Australian regional community adversely affected by winter wood smoke pollution. Our results indicate that residents with stronger positive affective associations with wood heating expressed less support for wood smoke mitigation policies involving regulation. This relationship was fully mediated by expected benefits and costs associated with wood heating. Affective associations were unrelated to public support for policies involving education and incentives, which were broadly endorsed by all segments of the community, and were more strongly associated with rational considerations. Latent profile analysis revealed no evidence to support the proposition that some community members experience internal “heart versus head” conflicts in which their positive affective associations with wood heating would be at odds with their risk judgments about the dangers of wood smoke pollution. Affective associations and cost/benefit judgments were very consistent with each other.

A methodological exploration to determine transportation disadvantage variables: the partial least square approach

Determining the key variables of transportation disadvantage remains a great challenge as the variables are commonly selected using ad-hoc techniques. In order to identify the variables, this research develops a transportation disadvantage framework by manipulating the capability approach. Developed framework is statistically analysed using partial least square-based software to determine the framework fitness. The statistical analysis identifies mobility and socioeconomic variables that significantly influence transportation disadvantage. The research reveals the key socioeconomic variables for transportation disadvantage in the case of Brisbane, Australia as household structure, presence of dependent family member, vehicle ownership, and driving licence possession.

Stefan problems for melting nanoscaled particles

Under certain conditions, the mathematical models governing the melting of nano-sized particles predict unphysical results, which suggests these models are incomplete. This thesis studies the addition of different physical effects to these models, using analytic and numerical techniques to obtain realistic and meaningful results. In particular, the mathematical "blow-up" of solutions to ill-posed Stefan problems is examined, and the regularisation of this blow-up via kinetic undercooling. Other effects such as surface tension, density change and size-dependent latent heat of fusion are also analysed.

Confirmatory factor analysis and invariance testing of the Young Carer of Parents Inventory (YCOPI)

Objective Research into youth caregiving in families where a parent experiences a significant medical condition has been hampered by a lack of contextually sensitive measures of the nature and breadth of young caregiving experiences. This study examined the factor structure and measurement invariance of such a measure called the Young Carer of Parents Inventory (YCOPI; Pakenham et al., 2006) using confirmatory factor analysis across 3 groups of youth. The YCOPI has 2 parts: YCOPI-A with 5 factors assessing caregiving experiences that are applicable to all caregiving contexts; YCOPI-B with 4 factors that tap dimensions related to youth caregiving in the context of parent illness. Design Two samples (ages 9–20 years) were recruited: a community sample of 2,429 youth from which 2 groups were derived (“healthy” family [HF], n = 1760; parental illness [PI], n = 446), and a sample of 130 youth of a parent with multiple sclerosis). Results With some modification, the YCOPI-A demonstrated a replicable factor structure across 3 groups, and exhibited only partial measurement invariance across the HF and PI groups. The impact of assuming full measurement invariance on latent mean differences appeared small, supporting use of the measure in research and applied settings when estimated using latent factors and controlling for measurement invariance. PI youth reported significantly higher scores than did HF youth on all YCOPI-A subscales. The YCOPI-B requires some modifications, and further development work is recommended. Conclusion The factor structure that emerged and the addition of new items constitutes the YCOPI-Revised. Findings support the use of the YCOPI-Revised in research and applied settings.

Doing whitefella research in blackfella communities in Australia : decolonizing method in sports related research

There is much for non-Indigenous researchers to consider when researching in Indigenous contexts. This paper is a story of discovery for two researchers working on a project with the Indigenous Sports Program section of the Australian Sports Commission. It documents the slow, meticulous and sometimes clumsy steps taken to gain access to communities and conduct research guided by a social justice ethic. The research was successful in that eventually it was possible to develop the trust of individuals and some of the Indigenous communities more broadly, so that information could be gathered and given within the context of shared understandings and mutual interest. However, it is the turbulent journey, filled as it is, with latent tendencies, privileged assumptions and eventually reflexive readings of the data, which remains the focus of this paper. Tentative recommendations are offered to those wishing to advance this politically and epistemologically challenging approach to culturally based research

Factors influencing the provision of end-of-life care in critical care settings : development and testing of a survey instrument

Aim. To develop and psychometrically test a survey instrument to identify the factors influencing the provision of end-of-life care by critical care nurses. Background. Following a decision to withdraw life-sustaining treatment, critical care nurses remain with the patient and their family providing end-of-life care. Identification of factors influencing the provision of this care can give evidence to inform practice development and support nurses. Design. A cross-sectional survey of critical care nurses. Method. An online survey was developed, reviewed by an expert panel and pilot tested to obtain preliminary evidence of its reliability and validity. In May 2011, a convenience sample of critical care nurses (n = 392, response rate 25%) completed the survey. The analytical approach to data obtained from the 58 items measured on a Likert scale included exploratory factor analysis and descriptive statistics. Results. Exploratory factor analysis identified eight factors influencing the provision of end-of-life care: emotional support for nurses, palliative values, patient and family preferences, resources, organizational support, care planning, knowledge and preparedness. Internal consistency of each latent construct was deemed satisfactory. The results of descriptive statistics revealed a strong commitment to the inclusion of families in end-of-life care and the value of this care in the critical care setting. Conclusion. This paper reports preliminary evidence of the psychometric properties of a new survey instrument. The findings may inform practice development opportunities to support critical care nurses in the provision of endof- life care and improve the care that patients and their families receive.

A multidimensional collaborative filtering fusion approach with dimensionality reduction

Multidimensional data are getting increasing attention from researchers for creating better recommender systems in recent years. Additional metadata provides algorithms with more details for better understanding the interaction between users and items. While neighbourhood-based Collaborative Filtering (CF) approaches and latent factor models tackle this task in various ways effectively, they only utilize different partial structures of data. In this paper, we seek to delve into different types of relations in data and to understand the interaction between users and items more holistically. We propose a generic multidimensional CF fusion approach for top-N item recommendations. The proposed approach is capable of incorporating not only localized relations of user-user and item-item but also latent interaction between all dimensions of the data. Experimental results show significant improvements by the proposed approach in terms of recommendation accuracy.

Tensor reduction for user profiling in personalized recommender systems

User profiling is the process of constructing user models which represent personal characteristics and preferences of customers. User profiles play a central role in many recommender systems. Recommender systems recommend items to users based on user profiles, in which the items can be any objects which the users are interested in, such as documents, web pages, books, movies, etc. In recent years, multidimensional data are getting more and more attention for creating better recommender systems from both academia and industry. Additional metadata provides algorithms with more details for better understanding the interactions between users and items. However, most of the existing user/item profiling techniques for multidimensional data analyze data through splitting the multidimensional relations, which causes information loss of the multidimensionality. In this paper, we propose a user profiling approach using a tensor reduction algorithm, which we will show is based on a Tucker2 model. The proposed profiling approach incorporates latent interactions between all dimensions into user profiles, which significantly benefits the quality of neighborhood formation. We further propose to integrate the profiling approach into neighborhoodbased collaborative filtering recommender algorithms. Experimental results show significant improvements in terms of recommendation accuracy.

An external field prior for the hidden Potts model with application to cone-beam computed tomography

In images with low contrast-to-noise ratio (CNR), the information gain from the observed pixel values can be insufficient to distinguish foreground objects. A Bayesian approach to this problem is to incorporate prior information about the objects into a statistical model. A method for representing spatial prior information as an external field in a hidden Potts model is introduced. This prior distribution over the latent pixel labels is a mixture of Gaussian fields, centred on the positions of the objects at a previous point in time. It is particularly applicable in longitudinal imaging studies, where the manual segmentation of one image can be used as a prior for automatic segmentation of subsequent images. The method is demonstrated by application to cone-beam computed tomography (CT), an imaging modality that exhibits distortions in pixel values due to X-ray scatter. The external field prior results in a substantial improvement in segmentation accuracy, reducing the mean pixel misclassification rate for an electron density phantom from 87% to 6%. The method is also applied to radiotherapy patient data, demonstrating how to derive the external field prior in a clinical context.

What determines weight loss behaviour and body weight satisfaction in Australia? Evidence from the HILDA survey

This thesis undertakes an empirical investigation to identify factors that influence the decision to undertake weight loss behaviour using the nationally representative HILDA dataset. Although many factors influenced the decision, the findings suggested that body weight satisfaction was the greatest determinant of weight loss dieting. This thesis therefore conducted a further empirical study to analyse the determinants of body weight satisfaction. A rank-hypothesis was found to better predict variation in body weight satisfaction levels than the absolute value of the individual's Body Mass Index (BMI) or the relative-norm hypothesis, which are commonly reported in the literature.

Refining user and item profiles based on multidimensional data for top-n item recommendation

In recommender systems based on multidimensional data, additional metadata provides algorithms with more information for better understanding the interaction between users and items. However, most of the profiling approaches in neighbourhood-based recommendation approaches for multidimensional data merely split or project the dimensional data and lack the consideration of latent interaction between the dimensions of the data. In this paper, we propose a novel user/item profiling approach for Collaborative Filtering (CF) item recommendation on multidimensional data. We further present incremental profiling method for updating the profiles. For item recommendation, we seek to delve into different types of relations in data to understand the interaction between users and items more fully, and propose three multidimensional CF recommendation approaches for top-N item recommendations based on the proposed user/item profiles. The proposed multidimensional CF approaches are capable of incorporating not only localized relations of user-user and/or item-item neighbourhoods but also latent interaction between all dimensions of the data. Experimental results show significant improvements in terms of recommendation accuracy.

Deriving word association networks from text corpora

This article presents and evaluates a model to automatically derive word association networks from text corpora. Two aspects were evaluated: To what degree can corpus-based word association networks (CANs) approximate human word association networks with respect to (1) their ability to quantitatively predict word associations and (2) their structural network characteristics. Word association networks are the basis of the human mental lexicon. However, extracting such networks from human subjects is laborious, time consuming and thus necessarily limited in relation to the breadth of human vocabulary. Automatic derivation of word associations from text corpora would address these limitations. In both evaluations corpus-based processing provided vector representations for words. These representations were then employed to derive CANs using two measures: (1) the well known cosine metric, which is a symmetric measure, and (2) a new asymmetric measure computed from orthogonal vector projections. For both evaluations, the full set of 4068 free association networks (FANs) from the University of South Florida word association norms were used as baseline human data. Two corpus based models were benchmarked for comparison: a latent topic model and latent semantic analysis (LSA). We observed that CANs constructed using the asymmetric measure were slightly less effective than the topic model in quantitatively predicting free associates, and slightly better than LSA. The structural networks analysis revealed that CANs do approximate the FANs to an encouraging degree.

A smooth transition logit model of the effects of deregulation in the electricity market

This paper introduces the smooth transition logit (STL) model that is designed to detect and model situations in which there is structural change in the behaviour underlying the latent index from which the binary dependent variable is constructed. The maximum likelihood estimators of the parameters of the model are derived along with their asymptotic properties, together with a Lagrange multiplier test of the null hypothesis of linearity in the underlying latent index. The development of the STL model is motivated by the desire to assess the impact of deregulation in the Queensland electricity market and ascertain whether increased competition has resulted in significant changes in the behaviour of the spot price of electricity, specifically with respect to the occurrence of periodic abnormally high prices. The model allows the timing of any change to be endogenously determined and also market participants' behaviour to change gradually over time. The main results provide clear evidence in support of a structural change in the nature of price events, and the endogenously determined timing of the change is consistent with the process of deregulation in Queensland.

Rapid detection of TNT in aqueous media by selective label free surface enhanced Raman spectroscopy

We report rapid and ultra-sensitive detection system for 2,4,6-trinitrotoluene (TNT) using unmodified gold nanoparticles and surface-enhanced Raman spectroscopy (SERS). First, Meisenheimer complex has been formed in aqueous solution between TNT and cysteamine in less than 15 min of mixing. The complex formation is confirmed by the development of a pink colour and a new UV–vis absorption band around 520 nm. Second, the developed Meisenheimer complex is spontaneously self-assembled onto unmodified gold nanoparticles through a stable Au–S bond between the cysteamine moiety and the gold surface. The developed mono layer of cysteamine-TNT is then screened by SERS to detect and quantify TNT. Our experimental results demonstrate that the SERS-based assay provide an ultra-sensitive approach for the detection of TNT down to 22.7 ng/L. The unambiguous fingerprint identification of TNT by SERS represents a key advantage for our proposed method. The new method provides high selectivity towards TNT over 2,4 DNT and picric acid. Therefore it satisfies the practical requirements for the rapid screening of TNT in real life samples where the interim 24-h average allowable concentration of TNT in waste water is 0.04 mg/L.