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.

Criminal profiling

One of the more widely recognized and practiced subspecialities within forensic criminology is that of criminal profiling. It has a long history, as detailed in Turvey (2008a). It also boasts a small library of distinct literature, with different methods and subspecialities all its own. Criminal profiling is a practice that has seen increasing popular and media attention over the past several decades. It has been depicted in popular fiction such as films like Silence of the Lambs (1991) and television programs like Criminal Minds (2005– present). It has also been applied in a number of high profile cases, including the “Washington Snipers” (see Turvey and McGrath, 2005, for an extended discussion of profiling and the media in the D.C. Sniper case). As a result, students of criminology commonly express an interest in studying criminal profiling with a view to becoming profilers themselves.

Criminal profiling : behavioural consistency, the homology assumption and case linkage

Criminal profiling is an investigative tool used around the world to infer the personality and behavioural characteristics of an offender based on their crime. Case linkage, the process of determining discreet connections between crimes of the same offender, is a practice that falls under the general banner of criminal profiling and has been widely criticized. Two theories, behavioural consistency and the homology assumption, are examined and their impact on profiling in general and case linkage specifically is discussed...

Optimisation algorithms for microarray biclustering

In providing simultaneous information on expression profiles for thousands of genes, microarray technologies have, in recent years, been largely used to investigate mechanisms of gene expression. Clustering and classification of such data can, indeed, highlight patterns and provide insight on biological processes. A common approach is to consider genes and samples of microarray datasets as nodes in a bipartite graphs, where edges are weighted e.g. based on the expression levels. In this paper, using a previously-evaluated weighting scheme, we focus on search algorithms and evaluate, in the context of biclustering, several variations of Genetic Algorithms. We also introduce a new heuristic “Propagate”, which consists in recursively evaluating neighbour solutions with one more or one less active conditions. The results obtained on three well-known datasets show that, for a given weighting scheme,optimal or near-optimal solutions can be identified.

Profiling unlicensed drivers and riders in Queensland. Report to Queensland Department of Transport and Main Roads

Unlicensed driving remains a serious problem for road safety, despite ongoing improvements in traffic law enforcement practices and technology. This report examines de-identified traffic infringement and sanction histories for drivers in Queensland who had lost their licence between 1st January 2003 and 31st December 2008. A total of 546,117 Queensland drivers were identified. Key areas discussed include the prevalence of unlicensed driving and the extent to which particular offences were detected amongst drivers with a licence sanction or disqualified licence.

Who are our repeating students? Profiling students persisting after failure

Research on attrition has focused on the economic significance of low graduation rates in terms of costs to students (fees that do not culminate in a credential) and impact on future income. For a student who fails a unit and repeats the unit multiple times, the financial impact is significant and lasting (Bexley, Daroesman, Arkoudis & James 2013). There are obvious advantages for the timely completion of a degree, both for the student and the institution. Advantages to students include fee minimisation, enhanced engagement opportunities, effectual pathway to employment and a sense of worth, morale and cohort-identity benefits. Work undertaken by the QUT Analytics Project in 2013 and 2014 explored student engagement patterns capturing a variety of data sources and specifically, the use of LMS amongst students in 804 undergraduate units in one semester. Units with high failure rates were given further attention and it was found that students who were repeating a unit were less likely to pass the unit than students attempting it for the first time. In this repeating cohort, academic and behavioural variables were consistently more significant in the modelling than were any demographic variables, indicating that a student’s performance at university is far more impacted by what they do once they arrive than it is by where they come from. The aim of this poster session is to examine the findings and commonalities of a number of case studies that articulated the engagement activities of repeating students (which included collating data from Individual Unit Reports, academic and peer advising programs and engagement with virtual learning resources). Understanding the profile of the repeating student cohort is therefore as important as considering the characteristics of successful students so that the institution might be better placed to target the repeating students and make proactive interventions as early as possible.

Importance of literature profiling: An archival analysis with illustrative examples for IS researchers

The importance of a thorough and systematic literature review has long been recognised across academic domains as critical to the foundation of new knowledge and theory evolution. Driven by an exponentially growing body of knowledge in the IS discipline, there has been a recent influx of guidance on how to conduct a literature review. As literature reviews are emerging as a standalone research method in itself, increasingly these method focused guidelines are of great interest, receiving acceptance at top tier IS publication outlets. Nevertheless, the finer details which offer justification for the selected content, and the effective presentation of supporting data has not been widely discussed in these method papers to date. This paper addresses this gap by exploring the concept of ‘literature profiling’ while arguing that it is a key aspect of a comprehensive literature review. The study establishes the importance of profiling for managing aspects such as quality assurance, transparency and the mitigation of selection bias. And then discusses how profiling can provide a valid basis for data analysis based on the attributes of selected literature. In essence, this study has conducted an archival analysis of literature (predominately from the IS domain) to present its main argument; the value for literature profiling, with supporting exemplary illustrations.

Architecture for service profiling

Service oriented architecture is gaining momentum. However, in order to be successful, the proper and up-to-date description of services is required. Such a description may be provided by service profiling mechanisms, such as one presented in this article. Service profile can be defined as an up-to-date description of a subset of non-functional properties of a service. It allows for service comparison on the basis of non-functional parameters, and choosing the service which is most suited to the needs of a user. In this article the notion of a service profile along with service profiling mechanism is presented as well as the architecture of a profiling system. © 2006 IEEE.

High-resolution SNP microarray investigation of copy number variations on chromosome 18 in a control cohort

Copy number variations (CNVs) as described in the healthy population are purported to contribute significantly to genetic heterogeneity. Recent studies have described CNVs using lymphoblastoid cell lines or by application of specifically developed algorithms to interrogate previously described data. However, the full extent of CNVs remains unclear. Using high-density SNP array, we have undertaken a comprehensive investigation of chromosome 18 for CNV discovery and characterisation of distribution and association with chromosome architecture. We identified 399 CNVs, of which loss represents 98%, 58% are less than 2.5 kb in size and 71% are intergenic. Intronic deletions account for the majority of copy number changes with gene involvement. Furthermore, one-third of CNVs do not have putative breakpoints within repetitive sequences. We conclude that replicative processes, mediated either by repetitive elements or microhomology, account for the majority of CNVs in the healthy population. Genomic instability involving the formation of a non-B structure is demonstrated in one region.

Genome-wide mapping of cystitis due to Streptococcus agalactiae and Escherichia coli in mice identifies a unique bladder transcriptome that signifies pathogen-specific antimicrobial defense against urinary tract infection

The most common causes of urinary tract infections (UTIs) are Gram-negative pathogens such as Escherichia coli; however, Gram-positive organisms including Streptococcus agalactiae, or group B streptococcus (GBS), also cause UTI. In GBS infection, UTI progresses to cystitis once the bacteria colonize bladder, but the host responses triggered in the bladder immediately following infection are largely unknown. Here, we used genome-wide expression profiling to map the bladder transcriptome of GBS UTI in mice infected transurethrally with uropathogenic GBS that was cultured from a 35 year-old women with cystitis. RNA from bladders was applied to Affymetrix Gene-1.0ST microarrays; qRT-PCR was used to analyze selected gene responses identified in array datasets. A surprisingly small significant gene list of 172 genes was identified at 24h; this compared to 2507 genes identified in a side-by-side comparison with uropathogenic E. coli (UPEC). No genes exhibited significantly altered expression at 2h in GBS-infected mice according to arrays despite high bladder bacterial loads at this early time point. The absence of a marked early host response to GBS juxtaposed with broad-based bladder responses activated by UPEC at 2h. Bioinformatics analyses including integrative systems-level network mapping revealed multiple activated biological pathways in the GBS cystitis transcriptome that regulate leukocyte activation, inflammation, apoptosis, and cytokine-chemokine biosynthesis. These findings define a novel, minimalistic type of bladder host response triggered by GBS UTI, which comprises collective antimicrobial pathways that differ dramatically from those activated by UPEC. Overall, this study emphasizes the unique nature of bladder immune activation mechanisms triggered by distinct uropathogens.

Genome-wide DNA methylation profiling of CD8+ T cells shows a distinct epigenetic signature to CD4+ T cells in multiple sclerosis patients

Background Multiple sclerosis (MS) is thought to be a T cell-mediated autoimmune disorder. MS pathogenesis is likely due to a genetic predisposition triggered by a variety of environmental factors. Epigenetics, particularly DNA methylation, provide a logical interface for environmental factors to influence the genome. In this study we aim to identify DNA methylation changes associated with MS in CD8+ T cells in 30 relapsing remitting MS patients and 28 healthy blood donors using Illumina 450K methylation arrays. Findings Seventy-nine differentially methylated CpGs were associated with MS. The methylation profile of CD8+ T cells was distinctive from our previously published data on CD4+ T cells in the same cohort. Most notably, there was no major CpG effect at the MS risk gene HLA-DRB1 locus in the CD8+ T cells. Conclusion CD8+ T cells and CD4+ T cells have distinct DNA methylation profiles. This case–control study highlights the importance of distinctive cell subtypes when investigating epigenetic changes in MS and other complex diseases.

Expression profiling of lymph nodes in tuberculosis patients reveal inflammatory milieu at site of infection

Extrapulmonary manifestations constitute 15 to 20% of tuberculosis cases, with lymph node tuberculosis (LNTB) as the most common form of infection. However, diagnosis and treatment advances are hindered by lack of understanding of LNTB biology. To identify host response, Mycobacterium tuberculosis infected lymph nodes from LNTB patients were studied by means of transcriptomics and quantitative proteomics analyses. The selected targets obtained by comparative analyses were validated by quantitative PCR and immunohistochemistry. This approach provided expression data for 8,728 transcripts and 102 proteins, differentially regulated in the infected human lymph node. Enhanced inflammation with upregulation of T-helper1-related genes, combined with marked dysregulation of matrix metalloproteinases, indicates tissue damage due to high immunoactivity at infected niche. This expression signature was accompanied by significant upregulation of an immunoregulatory gene, leukotriene A4 hydrolase, at both transcript and protein levels. Comparative transcriptional analyses revealed LNTB-specific perturbations. In contrast to pulmonary TB-associated increase in lipid metabolism, genes involved in fatty-acid metabolism were found to be downregulated in LNTB suggesting differential lipid metabolic signature. This study investigates the tissue molecular signature of LNTB patients for the first time and presents findings that indicate the possible mechanism of disease pathology through dysregulation of inflammatory and tissue-repair processes.

Clinician perspective on molecular profiling of non-small-cell lung cancer

The article by Meric-Bernstam et al1 that was recently published in Journal of Clinical Oncology raises important questions about the clinical application of large-scale genomic testing. We congratulate the authors for this ambitious study, which successfully profiled 2,000 consecutive patients with advanced cancer. The next-generation sequencing (NGS) platform was used for 1,749 of 2,000 patients (87.5%). Of 789 patients with potentially actionable mutations, 83 (11%, or 4% of screened population) were enrolled in a genomically matched clinical study. As the editorial2 accompanying the article by Meric-Bernstam et al1 pointed out, the 4% figure, albeit disappointing, may be an underestimate because cancers such as lung adenocarcinoma and melanoma, for which ≥ 50% of patients have actionable mutations, were under-represented. ...