Evaluating automatic road detection across a large aerial imagery collection

The automated extraction of roads from aerial imagery can be of value for tasks including mapping, surveillance and change detection. Unfortunately, there are no public databases or standard evaluation protocols for evaluating these techniques. Many techniques are further hindered by a reliance on manual initialisation, making large scale application of the techniques impractical. In this paper, we present a public database and evaluation protocol for the evaluation of road extraction algorithms, and propose an improved automatic seed finding technique to initialise road extraction, based on a combination of geometric and colour features.

A semantics-based approach to querying process model repositories

As business process management technology matures, organisations acquire more and more business process models. The resulting collections can consist of hundreds, even thousands of models and their management poses real challenges. One of these challenges concerns model retrieval where support should be provided for the formulation and efficient execution of business process model queries. As queries based on only structural information cannot deal with all querying requirements in practice, there should be support for queries that require knowledge of process model semantics. In this paper we formally define a process model query language that is based on semantic relationships between tasks. This query language is independent of the particular process modelling notation used, but we will demonstrate how it can be used in the context of Petri nets by showing how the semantic relationships can be determined for these nets in such a way that state space explosion is avoided as much as possible. An experiment with three large process model repositories shows that queries expressed in our language can be evaluated efficiently.

Towards semantic search and inference in electronic medical records : an approach using concept-based information retrieval

Background This paper presents a novel approach to searching electronic medical records that is based on concept matching rather than keyword matching. Aim The concept-based approach is intended to overcome specific challenges we identified in searching medical records. Method Queries and documents were transformed from their term-based originals into medical concepts as defined by the SNOMED-CT ontology. Results Evaluation on a real-world collection of medical records showed our concept-based approach outperformed a keyword baseline by 25% in Mean Average Precision. Conclusion The concept-based approach provides a framework for further development of inference based search systems for dealing with medical data.

The Working After Cancer Study (WACS): a population-based study of middle-aged workers diagnosed with colorectal cancer and their return to work experiences

Background The number of middle-aged working individuals being diagnosed with cancer is increasing and so too will disruptions to their employment. The aim of the Working After Cancer Study is to examine the changes to work participation in the 12 months following a diagnosis of primary colorectal cancer. The study will identify barriers to work resumption, describe limitations on workforce participation, and evaluate the influence of these factors on health-related quality of life. Methods/Design An observational population-based study has been designed involving 260 adults newly-diagnosed with colorectal cancer between January 2010 and September 2011 and who were in paid employment at the time they were diagnosed. These cancer cases will be compared to a nationally representative comparison group of 520 adults with no history of cancer from the general population. Eligible cases will have a histologically confirmed diagnosis of colorectal cancer and will be identified through the Queensland Cancer Registry. Data on the comparison group will be drawn from the Household, Income and Labour Dynamics in Australia (HILDA) Survey. Data collection for the cancer group will occur at 6 and 12 months after diagnosis, with work questions also asked about the time of diagnosis, while retrospective data on the comparison group will be come from HILDA Waves 2009 and 2010. Using validated instruments administered via telephone and postal surveys, data will be collected on socio-demographic factors, work status and circumstances, and health-related quality of life (HRQoL) for both groups while the cases will have additional data collected on cancer treatment and symptoms, work productivity and cancer-related HRQoL. Primary outcomes include change in work participation at 12 months, time to work re-entry, work limitations and change in HRQoL status. Discussion This study will address the reasons for work cessation after cancer, the mechanisms people use to remain working and existing workplace support structures and the implications for individuals, families and workplaces. It may also provide key information for governments on productivity losses.

Maintenance test section survey : data collection and analysis

A Maintenance Test Section Survey (MTSS) was conducted as part of a Peer State Review of the Texas Maintenance Program conducted October 5–7, 2010. The purpose of the MTSS was to conduct a field review of 34 highway test sections and obtain participants’ opinions about pavement, roadside, and maintenance conditions. The goal was to cross reference or benchmark TxDOT’s maintenance practices based on practices used by selected peer states. Representatives from six peer states (California, Georgia, Kansas, Missouri, North Carolina, and Washington) were invited to Austin to attend a 3-day Peer State Review of TxDOT Maintenance Practices Workshop and to participate in a field survey of a number of pre-selected one-mile roadway sections. It should be emphasized that the objective of the survey was not to evaluate and grade or score TxDOT’s road network but rather to determine whether the selected roadway sections met acceptable standards of service as perceived by Directors of Maintenance or senior maintenance managers from the peer states...

Health complaint commissions in Australia : time for a national approach to data collection

Health complaint statistics are important for identifying problems and bringing about improvements to health care provided by health service providers and to the wider health care system. This paper overviews complaints handling by the eight Australian state and territory health complaint entities, based on an analysis of data from their annual reports. The analysis shows considerable variation between jurisdictions in the ways complaint data are defined, collected and recorded. Complaints from the public are an important accountability mechanism and open a window on service quality. The lack of a national approach leads to fragmentation of complaint data and a lost opportunity to use national data to assist policy development and identify the main areas causing consumers to complain. We need a national approach to complaints data collection in order to better respond to patients’ concerns.

TiltZoom : tilt-based zooming control for easy one-handed mobile interactions

In this paper, we present TiltZoom, a collection of tilt-based interaction techniques designed for easy one-handed zooming on mobile devices. TiltZoom represents novel gestural interaction techniques, implemented using rate-of-rotation readings from a gyroscope, a sensor commonly embedded on current generation smart phones. We designed and experimented three variants of TiltZoom - Tilt Level, Tilt and Hold and Flip Gesture. The design decisions for all three variants are discussed in this paper and their performance, as well as subjective user experience are evaluated and compared against conventional touch-based zooming techniques. TiltZoom appears to be a worthy addition to current established collection of gesture-based mobile interaction techniques for zooming controls, especially when user has only one hand available when moving about.

Family-based programmes for preventing smoking by children and adolescents

BACKGROUND: There is evidence that children's decisions to smoke are influenced by family and friends. OBJECTIVES: To assess the effectiveness of interventions to help family members to strengthen non-smoking attitudes and promote non-smoking by children and other family members. SEARCH STRATEGY: We searched 14 electronic bibliographic databases, including the Cochrane Tobacco Addiction Group specialized register, MEDLINE, EMBASE, PsycINFO and CINAHL. We also searched unpublished material, and the reference lists of key articles. We performed both free-text Internet searches and targeted searches of appropriate websites, and we hand-searched key journals not available electronically. We also consulted authors and experts in the field. The most recent search was performed in July 2006. SELECTION CRITERIA: Randomized controlled trials (RCTs) of interventions with children (aged 5-12) or adolescents (aged 13-18) and family members to deter the use of tobacco. The primary outcome was the effect of the intervention on the smoking status of children who reported no use of tobacco at baseline. Included trials had to report outcomes measured at least six months from the start of the intervention. DATA COLLECTION AND ANALYSIS: We reviewed all potentially relevant citations and retrieved the full text to determine whether the study was an RCT and matched our inclusion criteria. Two authors independently extracted study data and assessed them for methodological quality. The studies were too limited in number and quality to undertake a formal meta-analysis, and we present a narrative synthesis. MAIN RESULTS: We identified 19 RCTs of family interventions to prevent smoking. We identified five RCTs in Category 1 (minimal risk of bias on all counts); nine in Category 2 (a risk of bias in one or more areas); and five in Category 3 (risks of bias in design and execution such that reliable conclusions cannot be drawn from the study).Considering the fourteen Category 1 and 2 studies together: (1) four of the nine that tested a family intervention against a control group had significant positive effects, but one showed significant negative effects; (2) one of the five RCTs that tested a family intervention against a school intervention had significant positive effects; (3) none of the six that compared the incremental effects of a family plus a school programme to a school programme alone had significant positive effects; (4) the one RCT that tested a family tobacco intervention against a family non-tobacco safety intervention showed no effects; and (5) the one trial that used general risk reduction interventions found the group which received the parent and teen interventions had less smoking than the one that received only the teen intervention (there was no tobacco intervention but tobacco outcomes were measured). For the included trials the amount of implementer training and the fidelity of implementation are related to positive outcomes, but the number of sessions is not. AUTHORS' CONCLUSIONS: Some well-executed RCTs show family interventions may prevent adolescent smoking, but RCTs which were less well executed had mostly neutral or negative results. There is thus a need for well-designed and executed RCTs in this area.

WebPut : Efficient web-based data imputation

In this paper, we present WebPut, a prototype system that adopts a novel web-based approach to the data imputation problem. Towards this, Webput utilizes the available information in an incomplete database in conjunction with the data consistency principle. Moreover, WebPut extends effective Information Extraction (IE) methods for the purpose of formulating web search queries that are capable of effectively retrieving missing values with high accuracy. WebPut employs a confidence-based scheme that efficiently leverages our suite of data imputation queries to automatically select the most effective imputation query for each missing value. A greedy iterative algorithm is also proposed to schedule the imputation order of the different missing values in a database, and in turn the issuing of their corresponding imputation queries, for improving the accuracy and efficiency of WebPut. Experiments based on several real-world data collections demonstrate that WebPut outperforms existing approaches.

Graph-based concept weighting for medical information retrieval

This paper presents a graph-based method to weight medical concepts in documents for the purposes of information retrieval. Medical concepts are extracted from free-text documents using a state-of-the-art technique that maps n-grams to concepts from the SNOMED CT medical ontology. In our graph-based concept representation, concepts are vertices in a graph built from a document, edges represent associations between concepts. This representation naturally captures dependencies between concepts, an important requirement for interpreting medical text, and a feature lacking in bag-of-words representations. We apply existing graph-based term weighting methods to weight medical concepts. Using concepts rather than terms addresses vocabulary mismatch as well as encapsulates terms belonging to a single medical entity into a single concept. In addition, we further extend previous graph-based approaches by injecting domain knowledge that estimates the importance of a concept within the global medical domain. Retrieval experiments on the TREC Medical Records collection show our method outperforms both term and concept baselines. More generally, this work provides a means of integrating background knowledge contained in medical ontologies into data-driven information retrieval approaches.

Uncertainty analysis of pollutant build-up modelling based on a Bayesian Weighted Least Squares approach

Reliable pollutant build-up prediction plays a critical role in the accuracy of urban stormwater quality modelling outcomes. However, water quality data collection is resource demanding compared to streamflow data monitoring, where a greater quantity of data is generally available. Consequently, available water quality data sets span only relatively short time scales unlike water quantity data. Therefore, the ability to take due consideration of the variability associated with pollutant processes and natural phenomena is constrained. This in turn gives rise to uncertainty in the modelling outcomes as research has shown that pollutant loadings on catchment surfaces and rainfall within an area can vary considerably over space and time scales. Therefore, the assessment of model uncertainty is an essential element of informed decision making in urban stormwater management. This paper presents the application of a range of regression approaches such as ordinary least squares regression, weighted least squares Regression and Bayesian Weighted Least Squares Regression for the estimation of uncertainty associated with pollutant build-up prediction using limited data sets. The study outcomes confirmed that the use of ordinary least squares regression with fixed model inputs and limited observational data may not provide realistic estimates. The stochastic nature of the dependent and independent variables need to be taken into consideration in pollutant build-up prediction. It was found that the use of the Bayesian approach along with the Monte Carlo simulation technique provides a powerful tool, which attempts to make the best use of the available knowledge in the prediction and thereby presents a practical solution to counteract the limitations which are otherwise imposed on water quality modelling.

Disaster strikes, then what? Using evaluation in narrative driven (oral history & digital storytelling) community-based projects

In 2011 Queensland suffered both floods and cyclones, leaving residents without homes and their communities in ruins (2011). This paper presents how researchers from QUT, who are also members of the Oral History Association of Australia (OHAA) Queensland’s chapter, are using oral history, photographs, videography and digital storytelling to help heal and empower rural communities around the state and how evaluation has become a key element of our research. QUT researchers ran storytelling workshops in the capital city of Brisbane i early 2011, after the city suffered sever flooding. Cyclone Yasi then struck the town of Cardwell (in February 2011) destroying their historical museum and recording equipment. We delivered an 'emergency workshop', offering participants hands on use of the equipment, ethical and interviewing theory, so that the community could start to build a new collection. We included oral history workshops as well as sessions on how best to use a video camera, digital camera and creative writing sessions, so the community would also know how to make 'products' or exhibition pieces out of the interviews they were recording. We returned six months later to conduct follow-up workshops and the material produced by and with the community had been amazing. More funding has now been secured to replicate audio/visual/writing workshops in other remote rural Queensland communities including Townsville, Mackay and Cunnamulla and Toowoomba in 2012, highlighting the need for a multi media approach, to leverage the most out of OH interviews as a mechanism to restore and promote community resilience and pride.

Framing local stories for community consumption facilitation and evaluation in narrative driven arts-based projects

Cyclone Yasi struck the Cassowary Coast of Queensland in the early hours of Feb 3, 2011, destroying many homes sand property, including the destruction of the Cardwell and district historical society’s premises. With their own homes flattened, many were forced to live in mobile accommodation, with extended family, or leave altogether. The historical society members however were more devastated by their flattened foreshore museum and loss of their collection material. A call for assistance was made through the OHAA Qld branch, who along with QUT sponsored a trip to somehow plan how they could start to pick up the pieces to start again. This presentation highlights the need for communities to gather, preserve and present their own stories, in a way that is sustainable and meaningful to them, but that good advice and support along the way is important. Two 2 day workshops were held in March and then September, augmented by plenty of email correspondence and phone calls in between. Participants learnt that if they could conduct quality oral history interviews, they could later use these in many exhibitable ways including: documentary pieces; digital stories; photographic collections; creative short stories; audio segments –while also drawing closely together a suffering community. This story is not only about the people who were interviewed about the night Yasi struck, but the amazing women (all over 50) of the historical society who were willing to try and leap the digital divide that faces older Australians, especially those in rural Australia, so that their older local stories would not be lost and so that new stories could also be remembered.

Visioning technology-based simulated learning environments in the social work curriculum

This report, prepared by QUT Social Work and Human Services, in collaboration with key sector stakeholders, identifies the current and potentially expanded uses of Simulated Learning Environments (SLE) as part of the Health Workforce Australia (HWA) National Project. An expert Reference Group guided the project, facilitated the data collection, and provided feedback and support on the findings and broad recommendations.

Heteroscedastic probabilistic linear discriminant analysis for manifold learning in video-based face recognition

To recognize faces in video, face appearances have been widely modeled as piece-wise local linear models which linearly approximate the smooth yet non-linear low dimensional face appearance manifolds. The choice of representations of the local models is crucial. Most of the existing methods learn each local model individually meaning that they only anticipate variations within each class. In this work, we propose to represent local models as Gaussian distributions which are learned simultaneously using the heteroscedastic probabilistic linear discriminant analysis (PLDA). Each gallery video is therefore represented as a collection of such distributions. With the PLDA, not only the within-class variations are estimated during the training, the separability between classes is also maximized leading to an improved discrimination. The heteroscedastic PLDA itself is adapted from the standard PLDA to approximate face appearance manifolds more accurately. Instead of assuming a single global within-class covariance, the heteroscedastic PLDA learns different within-class covariances specific to each local model. In the recognition phase, a probe video is matched against gallery samples through the fusion of point-to-model distances. Experiments on the Honda and MoBo datasets have shown the merit of the proposed method which achieves better performance than the state-of-the-art technique.