Clustering and labeling a web scale document collection using Wikipedia clusters

Clustering is an important technique in organising and categorising web scale documents. The main challenges faced in clustering the billions of documents available on the web are the processing power required and the sheer size of the datasets available. More importantly, it is nigh impossible to generate the labels for a general web document collection containing billions of documents and a vast taxonomy of topics. However, document clusters are most commonly evaluated by comparison to a ground truth set of labels for documents. This paper presents a clustering and labeling solution where the Wikipedia is clustered and hundreds of millions of web documents in ClueWeb12 are mapped on to those clusters. This solution is based on the assumption that the Wikipedia contains such a wide range of diverse topics that it represents a small scale web. We found that it was possible to perform the web scale document clustering and labeling process on one desktop computer under a couple of days for the Wikipedia clustering solution containing about 1000 clusters. It takes longer to execute a solution with finer granularity clusters such as 10,000 or 50,000. These results were evaluated using a set of external data.

Classification of inventory data suit for managing Australian railway bridges

With a focus to optimising the life cycle performance of Australian Railway bridges, new bridge classification and environmental classification systems are proposed. The new bridge classification system is mainly to facilitate the implementation of novel Bridge Management System (BMS) which optimise the life cycle cost both at project level and network level while environment classification is mainly to improve accuracy of Remaining Service Potential (RSP) module of the proposed BMS. In fact, limited capacity of the existing BMS to trigger the maintenance intervention point is an indirect result of inadequacies of the existing bridge and environmental classification systems. The proposed bridge classification system permits to identify the intervention points based on percentage deterioration of individual elements and maintenance cost, while allowing performance based rating technique to implement for maintenance optimisation and prioritisation. Simultaneously, the proposed environment classification system will enhance the accuracy of prediction of deterioration of steel components.

Using viewer’s facial expression and heart rate for sports video highlights detection

Viewer interests, evoked by video content, can potentially identify the highlights of the video. This paper explores the use of facial expressions (FE) and heart rate (HR) of viewers captured using camera and non-strapped sensor for identifying interesting video segments. The data from ten subjects with three videos showed that these signals are viewer dependent and not synchronized with the video contents. To address this issue, new algorithms are proposed to effectively combine FE and HR signals for identifying the time when viewer interest is potentially high. The results show that, compared with subjective annotation and match report highlights, ‘non-neutral’ FE and ‘relatively higher and faster’ HR is able to capture 60%-80% of goal, foul, and shot-on-goal soccer video events. FE is found to be more indicative than HR of viewer’s interests, but the fusion of these two modalities outperforms each of them.

Playing Pranks: Traditions, communities & co-creative cultural performance practice

Pranks, hoaxes and practical jokes are co-creative cultural performance practices that appear across times, contexts and cultures. These practices include everyday play amongst families, friends and coworkers, entertainment programs such as Prank Patrol, Punked or Scare Tactics, and aesthetic and activist pranks perpetrated by situationist artists, guerrilla artists, and, most recently, culture ‘jammers’ or ‘hackers’ intent on turning capitalist systems back on themselves. Although it can, in common usage, describe almost any show off behaviour, a prank in the strictest definition of the term is a performance that deploys a very specific set of strategies. It is an act of trickery, mischief, or deceit, that must be taken as real, and momentarily cause real fear, anger or worry for an unwitting spectator-become-performer, who is meant to play along until the trick is revealed and their response can be represented back to the prankster, other spectators, or society as a whole, either for the sake of entertainment or for the sake of commentary on a cultural phenomenon. A prank, in this sense, deliberately blurs the boundaries between daily and dramatic performance. It creates a moment of uncertainty, in which both the prankster’s ability to be creative, clever, or culturally astute, and the prankee’s ability to play along, discern the trick, discern the point of the trick, and, in the end, be duped, be a good sport, or even play/pay the prankster back, are both put to the test. In this paper, I consider a number of pranking traditions popular where I am in Australia, from the community-building pranks of footballers, bucks parties and ‘drop bear’ tales told to tourists, to the more controversial pranks of radio shock jocks, activists and artists. I use performance, spectatorship and ethical theory to examine the engagement between prankster, pranked spectator, and other spectators, in this most distinctive sort of community-driven performance practice, and the way it builds and breaks status, social and other sorts of relationships within and between specific communities.

Developing creative services industries in Hangzhou, China

Creativity is changing the People’s Republic of China according to Li Wuwei (2011), a leading Chinese economist and policy advisor. The nation is learning to embrace a “third industrial revolution” (Rifkin, 2011) while banking the economic capital of the carbon-dependent manufacturing economy. Urbanisation is also driving change and consumer culture (Gerth, 2010). Most of China’s high-value creative service industries are found in the large urban centres of Beijing, Shanghai, Guangzhou and Shenzhen in the coastal provinces. China’s second-tier cities, including Hangzhou in Zhejiang province, are also seeking to make capital out of culture, albeit with different strategies than the coastal hubs. The Hangzhou metropolitan area is the fourth largest in China, with 8.8 million residents. Zhejiang province was once known as the “land of rice and fish.” However, with the increased emphasis on productivity in China’s economic reforms since 1978, the province became an economic heavyweight, characterised by small and medium-sized enterprises often working together to produce complementary products...

Centroid training to achieve effective text classification

Traditional text classification technology based on machine learning and data mining techniques has made a big progress. However, it is still a big problem on how to draw an exact decision boundary between relevant and irrelevant objects in binary classification due to much uncertainty produced in the process of the traditional algorithms. The proposed model CTTC (Centroid Training for Text Classification) aims to build an uncertainty boundary to absorb as many indeterminate objects as possible so as to elevate the certainty of the relevant and irrelevant groups through the centroid clustering and training process. The clustering starts from the two training subsets labelled as relevant or irrelevant respectively to create two principal centroid vectors by which all the training samples are further separated into three groups: POS, NEG and BND, with all the indeterminate objects absorbed into the uncertain decision boundary BND. Two pairs of centroid vectors are proposed to be trained and optimized through the subsequent iterative multi-learning process, all of which are proposed to collaboratively help predict the polarities of the incoming objects thereafter. For the assessment of the proposed model, F1 and Accuracy have been chosen as the key evaluation measures. We stress the F1 measure because it can display the overall performance improvement of the final classifier better than Accuracy. A large number of experiments have been completed using the proposed model on the Reuters Corpus Volume 1 (RCV1) which is important standard dataset in the field. The experiment results show that the proposed model has significantly improved the binary text classification performance in both F1 and Accuracy compared with three other influential baseline models.

Medical semantic similarity with a neural language model

Advances in neural network language models have demonstrated that these models can effectively learn representations of words meaning. In this paper, we explore a variation of neural language models that can learn on concepts taken from structured ontologies and extracted from free-text, rather than directly from terms in free-text. This model is employed for the task of measuring semantic similarity between medical concepts, a task that is central to a number of techniques in medical informatics and information retrieval. The model is built with two medical corpora (journal abstracts and patient records) and empirically validated on two ground-truth datasets of human-judged concept pairs assessed by medical professionals. Empirically, our approach correlates closely with expert human assessors ($\approx$ 0.9) and outperforms a number of state-of-the-art benchmarks for medical semantic similarity. The demonstrated superiority of this model for providing an effective semantic similarity measure is promising in that this may translate into effectiveness gains for techniques in medical information retrieval and medical informatics (e.g., query expansion and literature-based discovery).

Exploiting inference from semantic annotations for information retrieval

The increasing amount of information that is annotated against standardised semantic resources offers opportunities to incorporate sophisticated levels of reasoning, or inference, into the retrieval process. In this position paper, we reflect on the need to incorporate semantic inference into retrieval (in particular for medical information retrieval) as well as previous attempts that have been made so far with mixed success. Medical information retrieval is a fertile ground for testing inference mechanisms to augment retrieval. The medical domain offers a plethora of carefully curated, structured, semantic resources, along with well established entity extraction and linking tools, and search topics that intuitively require a number of different inferential processes (e.g., conceptual similarity, conceptual implication, etc.). We argue that integrating semantic inference in information retrieval has the potential to uncover a large amount of information that otherwise would be inaccessible; but inference is also risky and, if not used cautiously, can harm retrieval.

The sweet aroma of success: Modification of products for new markets in the case of TORQ by Di Bella Coffee

In 2002, Phillip Di Bella’s childhood passion for coffee and keen entrepreneurial spirit led him to establish a small coffee roasting warehouse in in the inner suburbs of Brisbane (Di Bella, 2012). With a keen sense of direction and passion for his coffee products and providing unparalleled customer service, Di Bella Coffee quickly grew to become a key player in the coffee roasting scene. This passion for the ultimate coffee experience is evident in the firm’s logo ‘Di Bella Coffee Inspires Passion’. Phillip Di Bella stated that ‘the common denominator of this company is about inspiration and passion. We are not a coffee company, we are a people company. You know, are we inspiring you from the moment you walk in the door to the moment you leave. If you are not feeling inspired then we haven’t done our job properly as a company’. Fundamentally, providing the ultimate coffee experience, as detailed in the following case is one in which focuses on the coffee consumption experience, not the coffee itself. Over that last 10 years Di Bella Coffee has constantly strived for the ultimate coffee, while expanding business operations into the booming Asian coffee market, establishing headquarters in Shanghai in 2010. In 2011, Di Bella Coffee commenced their second international venture with the launch of operations in India (Di Bella Coffee, 2012); followed shortly by the creation of a new category of coffee, set to revolutionise to coffee industry. The fusion of two traditional forms of coffee; espresso coffee and instant coffee, to create a third category- espresso instant, led to the development of TORQ by Di Bella.

Association of STAT3 and TNFRSF1A with ankylosing spondylitis in Han Chinese

Objectives: Recent association studies by the Australo-Anglo-American Spondyloarthritis Consortium (TASC) in Caucasian European populations from Australia, North America and the UK have identified a number of genes as being associated with ankylosing spondylitis (AS). A candidate gene study in a Han Chinese population was performed based on these findings to identify associated genes in this population. Methods: A case-control study was performed in a Han Chinese population of patients with AS (n=775) and controls (n=1587) from Shanghai and Nanjing. All patients met the modified New York criteria for AS. The cases and controls were genotyped for 115 single nucleotide polymorphisms (SNPs) tagging IL23R, ERAP1, STAT3, JAK2, TNFRSF1A and TRADD, as well as other confirmation SNPs from the TASC study, using the Sequenom iPlex and the ABI OpenArray platforms. Statistical analysis of genotyped SNPs was performed using the Cochran - Armitage test for trend and meta-analysis was performed using METAL. SNPs in AS-associated genes in this study were then imputed using MaCH, and association with AS tested by logistic regression. Results: SNPs in TNFRSF1A (rs4149577, p=8.2×10-4), STAT3 (rs2293152, p=0.0015; rs1053005, p=0.017) and ERAP1 (rs27038, p=0.0091; rs27037, p=0.0092) were significantly associated with AS in Han Chinese. Association was also observed between AS and the intergenic region 2p15 (rs10865331, p=0.023). The lack of association between AS and IL23R in Han Chinese was confirmed (all SNPs p>0.1). Conclusions: The study results demonstrate for the first time that genetic polymorphisms in STAT3, TNFRSF1A and 2p15 are associated with AS in Han Chinese, suggesting common pathogenic mechanisms for the disease in Chinese and Caucasian European populations. Furthermore, previous findings demonstrating that ERAP1, but not IL23R, is associated with AS in Chinese patients were confirmed.

Association of ERAP1, but not IL23R, with ankylosing spondylitis in a Han Chinese population

Objective The results of a recent genome-wide association study have shown that ERAP1 and IL23R are associated with ankylosing spondylitis (AS) in Caucasian populations from North America and the UK. Based on these findings, we undertook the current study to investigate whether single-nucleotide polymorphisms (SNPs) covering the genes ERAP1 and IL23R are associated with AS in a Han Chinese population. Methods A case-control study was performed in Han Chinese patients with AS (n = 527) and controls (n = 945) from Shanghai and Nanjing. All patients met the modified New York criteria for AS. The Sequenom iPlex platform was used to genotype cases and controls for 21 tag SNPs covering IL23R and 38 tag SNPs covering ERAP1. Statistical analysis was performed using the Cochran-Armitage test for trend. Results Multiple SNPs in ERAP1 were significantly associated with AS (for rs27980, P = 0.0048; for rs7711564, P = 0.0081). However, no association was observed between IL23R and AS (for all SNPs, P > 0.1). The nonsynonymous SNP in IL23R, rs11209026, widely thought to be the primary AS-associated SNP in IL23R in Europeans, was found not to be polymorphic in Chinese. Conclusion Our results demonstrate that genetic polymorphisms in ERAP1 are associated with AS in Han Chinese, suggesting a common pathogenic mechanism for the disease in Chinese and Caucasian populations, and that IL23R is not associated with AS in Chinese, indicating a difference in the mechanism of disease pathogenesis between Chinese and Caucasian populations. This may result from the fact that rs11209026, the nonsynonymous SNP in IL23R, is not polymorphic in Chinese patients, providing further evidence that rs11209026 is the key polymorphism associated with AS (and likely inflammatory bowel disease and psoriasis) in this gene.

Adherence to insulin treatment in diabetes: Can it be improved?

Insulin is used in all subjects with Type 1 diabetes, and when Type 2 diabetes is not controlled by oral anti-diabetic medicines, insulin is also used in Type 2 diabetes. However, despite this use, there is still increased mortality and morbidity in subjects with diabetes, compared to subjects without diabetes. One of the factors, which may be involved in this increased mortality and morbidity in subjects with diabetes, is nonadherence to insulin. Nonadherence rates to insulin are in the range20-38%, and many factors contribute to the nonadherence. The major aim of the review was to determine whether interventions to improve adherence to insulin do actually improve adherence to insulin. Most studies have shown that adherence to insulin was improved by changing from the vial-and-syringe approach to prefilled insulin pens, but not all studies have shown that this translated into better glycemic control and clinical outcomes. The results of studies using automatic telephone messages to improve adherence to insulin to date are inconclusive. There is limited and variable evidence that an intervention by a nurse/educator, which discusses adherence to medicines, does improve adherence to insulin. In contrast, there is little or no evidence that an extra intervention by a doctor or an intervention by a pharmacist, which discusses adherence to insulin, does actually improve the measured adherence to insulin. In conclusion, rather than assuming that an intervention by a health professional discussing adherence to insulin actually improves adherence to insulin, long-term studies investigating this are required.

Critical risk factors affecting the implementation of PPP waste-to-energy projects in China

Since 2003, Mainland China has been promoting the public–private partnership (PPP) procurement model in the waste-to-energy incineration sector to reduce the waste burying rate and improve environmental quality. Five critical risk factors (CRFs) that affect the construction and operation of waste-to-energy incineration projects have been identified from real-life risk events of 14 PPP waste-to-energy incineration plants through content analysis. These risk factors are insufficient waste supply, disposal of non-licensed waste, environmental risk, payment risk, and lack of supporting infrastructure. A recently completed PPP waste-to-energy incineration plant, the Shanghai Tianma project, was investigated to learn from the effective management of CRFs. First-hand data about the Shanghai Tianma project was collected, with a focus on project negotiation and concession agreement. Lessons learned about risk management were acquired. This paper presents a detailed study of the contractual structure, risk sharing scheme, risk response measures to CRFs, and project transfer of a PPP project. The study results will provide governments with management implications to prepare equitable concession agreements and benefit private investors by effectively mitigating and managing risks in future PPP waste-to-energy incineration projects.

Impact of temperature on mortality in three major Chinese cities

To study the relation between temperature and mortality by estimating the temperature-related mortality in Beijing, Shanghai, and Guangzhou. METHODS: Data of daily mortality, weather and air pollution in the three cities were collected. A distributed lag nonlinear model was established and used in analyzing the effects of temperature on mortality. Current and future net temperature-related mortality was estimated. RESULTS: The association between temperature and mortality was J-shaped, with an increased death risk of both hot and cold temperature in these cities. The effects of cold temperature on health lasted longer than those of hot temperature. The projected temperature-related mortality increased with the decreased cold-related mortality. The mortality was higher in Guangzhou than in Beijing and Shanghai. CONCLUSION: The impact of temperature on health varies in the 3 cities of China, which may have implications for climate policy making in China.

The effects of ambient temperature on cerebrovascular mortality: An epidemiologic study in four climatic zones in China

Background Little evidence is available about the association between temperature and cerebrovascular mortality in China. This study aims to examine the effects of ambient temperature on cerebrovascular mortality in different climatic zones in China. Method We obtained daily data on weather conditions, air pollution and cerebrovascular deaths from five cities (Beijing, Tianjin, Shanghai, Wuhan, and Guangzhou) in China during 2004-2008. We examined city-specific associations between ambient temperature and the cerebrovascular mortality, while adjusting for season, long-term trends, day of the week, relative humidity and air pollution. We examined cold effects using a 1°C decrease in temperature below a city-specific threshold, and hot effects using a 1°C increase in temperature above a city-specific threshold. We used a meta-analysis to summarize the cold and hot effects across the five cities. Results Beijing and Tianjin (with low mean temperature) had lower thresholds than Shanghai, Wuhan and Guangzhou (with high mean temperature). In Beijing, Tianjin, Wuhan and Guangzhou cold effects were delayed, while in Shanghai there was no or short induction. Hot effects were acute in all five cities. The cold effects lasted longer than hot effects. The hot effects were followed by mortality displacement. The pooled relative risk associated with a 1°C decrease in temperature below thresholds (cold effect) was 1.037 (95% confidence interval (CI): 1.020, 1.053). The pooled relative risk associated with a 1°C increase in temperature above thresholds (hot effect) was 1.014 (95% CI: 0.979, 1.050). Conclusion Cold temperatures are significantly associated with cerebrovascular mortality in China, while hot effect is not significant. People in colder climate cities were sensitive to hot temperatures, while people in warmer climate cities were vulnerable to cold temperature.