Adaptable, model-driven security engineering for SaaS cloud-based applications

Software-as-a-service (SaaS) multi-tenancy in cloud-based applications helps service providers to save cost, improve resource utilization, and reduce service customization and maintenance time. This is achieved by sharing of resources and service instances among multiple "tenants" of the cloud-hosted application. However, supporting multi-tenancy adds more complexity to SaaS applications required capabilities. Security is one of these key requirements that must be addressed when engineering multi-tenant SaaS applications. The sharing of resources among tenants - i.e. multi-tenancy - increases tenants' concerns about the security of their cloud-hosted assets. Compounding this, existing traditional security engineering approaches do not fit well with the multi-tenancy application model where tenants and their security requirements often emerge after the applications and services were first developed. The resultant applications do not usually support diverse security capabilities based on different tenants' needs, some of which may change at run-time i.e. after cloud application deployment. We introduce a novel model-driven security engineering approach for multi-tenant, cloud-hosted SaaS applications. Our approach is based on externalizing security from the underlying SaaS application, allowing both application/service and security to evolve at runtime. Multiple security sets can be enforced on the same application instance based on different tenants' security requirements. We use abstract models to capture service provider and multiple tenants' security requirements and then generate security integration and configurations at runtime. We use dependency injection and dynamic weaving via Aspect-Oriented Programming (AOP) to integrate security within critical application/service entities at runtime. We explain our approach, architecture and implementation details, discuss a usage example, and present an evaluation of our approach on a set of open source web applications.

Avian influenza infection dynamics under variable climatic conditions, viral prevalence is rainfall driven in waterfowl from temperate, south-east Australia

Understanding Avian Influenza Virus (AIV) infection dynamics in wildlife is crucial because of possible virus spill over to livestock and humans. Studies from the northern hemisphere have suggested several ecological and environmental drivers of AIV prevalence in wild birds. To determine if the same drivers apply in the southern hemisphere, where more irregular environmental conditions prevail, we investigated AIV prevalence in ducks in relation to biotic and abiotic factors in south-eastern Australia. We sampled duck faeces for AIV and tested for an effect of bird numbers, rainfall anomaly, temperature anomaly and long-term ENSO (El-Niño Southern Oscillation) patterns on AIV prevalence. We demonstrate a positive long term effect of ENSO-related rainfall on AIV prevalence. We also found a more immediate response to rainfall where AIV prevalence was positively related to rainfall in the preceding 3-7 months. Additionally, for one duck species we found a positive relationship between their numbers and AIV prevalence, while prevalence was negatively or not affected by duck numbers in the remaining four species studied. In Australia largely non-seasonal rainfall patterns determine breeding opportunities and thereby influence bird numbers. Based on our findings we suggest that rainfall influences age structures within populations, producing an influx of immunologically naïve juveniles within the population, which may subsequently affect AIV infection dynamics. Our study suggests that drivers of AIV dynamics in the northern hemisphere do not have the same influence at our south-east Australian field site in the southern hemisphere due to more erratic climatological conditions.

A model driven approach to care planning systems for consumer engagement in chronic disease management

This work demonstrates a model-driven approach to the development of care plan systems, amenable to: (a) a flexible and extensible definition of care plan scope; and (b) deployment of care plan viewing and tracking functionality to a wide range of physical computing devices. The approach utilises a care plan domain model from which guideline implementers formulate care plan templates aligning to specific clinical guidelines. A clinical end user would subsequently constrain that template (e.g., selecting a subset of available activities and specific targets) to create a care plan instance for an individual patient. An XML care plan visualisation definition created using the Marama tool is transformed to OpenLaszlo script from which Shockwave Flash objects can be compiled, creating Flash applications that run on a variety of hardware for both clinical and patient users. The approach is illustrated with respect to an overweight and obesity guideline.

Learning vector representation of medical objects via EMR-driven nonnegative restricted Boltzmann machines (eNRBM)

Electronic medical record (EMR) offers promises for novel analytics. However, manual feature engineering from EMR is labor intensive because EMR is complex - it contains temporal, mixed-type and multimodal data packed in irregular episodes. We present a computational framework to harness EMR with minimal human supervision via restricted Boltzmann machine (RBM). The framework derives a new representation of medical objects by embedding them in a low-dimensional vector space. This new representation facilitates algebraic and statistical manipulations such as projection onto 2D plane (thereby offering intuitive visualization), object grouping (hence enabling automated phenotyping), and risk stratification. To enhance model interpretability, we introduced two constraints into model parameters: (a) nonnegative coefficients, and (b) structural smoothness. These result in a novel model called eNRBM (EMR-driven nonnegative RBM). We demonstrate the capability of the eNRBM on a cohort of 7578 mental health patients under suicide risk assessment. The derived representation not only shows clinically meaningful feature grouping but also facilitates short-term risk stratification. The F-scores, 0.21 for moderate-risk and 0.36 for high-risk, are significantly higher than those obtained by clinicians and competitive with the results obtained by support vector machines.