Policy recognition in the Abstract Hidden Markov Model

In this paper, we present a method for recognising an agent's behaviour in dynamic, noisy, uncertain domains, and across multiple levels of abstraction. We term this problem on-line plan recognition under uncertainty and view it generally as probabilistic inference on the stochastic process representing the execution of the agent's plan. Our contributions in this paper are twofold. In terms of probabilistic inference, we introduce the Abstract Hidden Markov Model (AHMM), a novel type of stochastic processes, provide its dynamic Bayesian network (DBN) structure and analyse the properties of this network. We then describe an application of the Rao-Blackwellised Particle Filter to the AHMM which allows us to construct an efficient, hybrid inference method for this model. In terms of plan recognition, we propose a novel plan recognition framework based on the AHMM as the plan execution model. The Rao-Blackwellised hybrid inference for AHMM can take advantage of the independence properties inherent in a model of plan execution, leading to an algorithm for online probabilistic plan recognition that scales well with the number of levels in the plan hierarchy. This illustrates that while stochastic models for plan execution can be complex, they exhibit special structures which, if exploited, can lead to efficient plan recognition algorithms. We demonstrate the usefulness of the AHMM framework via a behaviour recognition system in a complex spatial environment using distributed video surveillance data.

The hidden permutation model and location-based activity recognition

Permutation modeling is challenging because of the combinatorial nature of the problem. However, such modeling is often required in many real-world applications, including activity recognition where subactivities are often permuted and partially ordered. This paper introduces a novel Hidden Permutation Model (HPM) that can learn the partial ordering constraints in permuted state sequences. The HPM is parameterized as an exponential family distribution and is flexible so that it can encode constraints via different feature functions. A chain-flipping Metropolis-Hastings Markov chain Monte Carlo (MCMC) is employed for inference to overcome the O(n!) complexity. Gradient-based maximum likelihood parameter learning is presented for two cases when the permutation is known and when it is hidden. The HPM is evaluated using both simulated and real data from a location-based activity recognition domain. Experimental results indicate that the HPM performs far better than other baseline models, including the naive Bayes classifier, the HMM classifier, and Kirshner's multinomial permutation model. Our presented HPM is generic and can potentially be utilized in any problem where the modeling of permuted states from noisy data is needed.

Learning and detecting activities from movement trajectories using the hierarchical hidden Markov model

Directly modeling the inherent hierarchy and shared structures of human behaviors, we present an application of the hierarchical hidden Markov model (HHMM) for the problem of activity recognition. We argue that to robustly model and recognize complex human activities, it is crucial to exploit both the natural hierarchical decomposition and shared semantics embedded in the movement trajectories. To this end, we propose the use of the HHMM, a rich stochastic model that has been recently extended to handle shared structures, for representing and recognizing a set of complex indoor activities. Furthermore, in the need of real-time recognition, we propose a Rao-Blackwellised particle filter (RBPF) that efficiently computes the filtering distribution at a constant time complexity for each new observation arrival. The main contributions of this paper lie in the application of the shared-structure HHMM, the estimation of the model's parameters at all levels simultaneously, and a construction of an RBPF approximate inference scheme. The experimental results in a real-world environment have confirmed our belief that directly modeling shared structures not only reduces computational cost, but also improves recognition accuracy when compared with the tree HHMM and the flat HMM.

A mutation-based evolving neural network model and its application to condition monitoring

Data analysis using intelligent systems is a key solution to many industrial problems. In this paper, a mutation-based evolving artificial neural network, which is based on an integration of the Fuzzy ARTMAP (FAM) neural network and evolutionary programming (EP), is proposed. The proposed FAMEP model is applied to detect and classify possible faults from a number of sensory signals of a circulating water system in a power generation plant. The efficiency of FAM-EP is assessed and compared with that of the original FAM network in terms of classification accuracy as well as network complexity. In addition, the bootstrap method is used to quantify the performance statistically. The results positively demonstrate the usefulness of FAM-EP in tackling data classification problems.

Multiple perspectives framework to model complex processes

The growing complexity of organisations has resulted in collaboration between multiple stakeholders becoming a challenging and critical issue that organisations must address in order to ensure their practices are sustainable. A multiple-case field study was conducted in order to demonstrate the proposed methodology of analysis and examination for knowledge-based systems in an actual organisational setting. The use of a multiple-perspective framework to improve understanding of the complex relationships in such systems was examined. In particular, the case study focused on the Australian Government’s Nation Building Economic Stimulus Plan (NBESP) which provided $1.9 billion to construct social housing across the State over two years. The results suggest that the use of a multi-perspective framework is appropriate and that there is a need for attention to be paid to the economic perspective.

Co-evolution path model : how enterprises as complex systems survive on the edge of chaos

In this theoretical paper, we introduce and describe a model, and demonstrate its origins from the disciplines of Enterprise Architecture, cybernetics and systems theory. We use cybernetic thinking to develop a ‘Co-evolution Path Model’ that describes how enterprises as complex systems co-evolve with their complex environments. The model re-interprets Stafford Beer’s Viable System Model, and also uses the theorem of the ‘good regulator’ of Conant and Ashby, exemplifying how various complexity management theories could be synthesised into a cybernetic theory of Enterprise Architecture, using concepts from the generalisation of EA frameworks.

Zebrafish as a genetic model in pre-clinical drug testing and screening

The traditional drug discovery pipeline for the identification and development of compounds that selectively target specific molecules to ameliorate disease remains a major focus for medical research. However, the zebrafish is increasingly providing alternative strategies for various components of this pipeline. Zebrafish and their embryos are small, easily accessible and relatively low cost, making them applicable to high-throughput, small molecule screening. Zebrafish can also be manipulated by a range of forward and reverse genetics techniques to facilitate gene discovery and functional studies. Moreover, their physiological and developmental complexity provides accurate models of human disease to underpin mechanism of action and in vivo validation studies. Finally, several of these biological characteristics make zebrafish eminently suitable for toxicity testing, including eco-toxicology. Here we review the application of zebrafish to preclinical drug development and toxicity testing, including recent advances in mutant generation, drug screening and toxicology that serve to further enhance the capabilities of this valuable model organism in drug discovery.

Developing and using an S3R model to analyze reasoning in web-based cross-national exchanges on sustainability

Within the increasing body of research that examines students' reasoning on socioscientific issues, we consider in particular student reasoning concerning acute, open-ended questions that bring out the complexities and uncertainties embedded in ill-structured problems. In this paper, we propose a socioscientific sustainability reasoning (S3R) model to analyze students' reasoning exchanges on environmental socially acute questions (ESAQs). The paper describes the development of an epistemological analysis of how sustainability perspectives can be integrated into socioscientific reasoning, which emphasizes the need for S3R to be both grounded in context and collective. We argue the complexity of ESAQs requires a consideration of multiple dimensions that form the basis of our S3R analysis model: problematization, interactions, knowledge, uncertainties, values, and governance. For each dimension, in the model we have identified indicators of four levels of complexity. We investigated the usefulness of the model in identifying improvements in reasoning that flow from cross-national web-based exchanges between groups of French and Australian students, concerning a local and a global ESAQ. The S3R model successfully captured the nature of reasoning about socioscientific sustainability issues, with the collective negotiation of multiple forms of knowledge as a key characteristic in improving reasoning levels. The paper provides examples of collaborative argumentation in collective texts (wikis) to illustrate the various levels of reasoning in each dimension, and diagrammatic representation of the evolution of collective reflections. We observe that a staged process of construction and confrontation, involving groups representing to some extent different cultural and contextual stances, is powerful in eliciting reasoned argument of enhanced quality. © 2014 Wiley Periodicals, Inc.

Mandatory jury directions in sexual assault trials in Victoria: less a model than a cautionary tale?

 Jury directions in relation to the issue of consent in trials of sexual offence cases are mandated in two jurisdictions in Australia (Victoria and the Northern Territory). The Australian Law Reform Commission, in conjunction with the New South Wales Law Reform Commission, has recommended that provisions similar to those in Victoria should be contained in relevant legislation in all States and Territories. However, a recent series of cases in Victoria has revealed significant problems in relation to the mandatory jury directions. These difficulties have generated increasingly elaborate and complex directions. The complexity of these directions not only presents considerable challenges for judges but also may overwhelm, rather than assist, members of the jury. The Court of Appeal of Victoria has called for “urgent and wholesale reform”. In the light of these concerns, it is suggested that the Victorian mandatory directions do not provide a model for other jurisdictions. Rather, the Victorian experience can be seen as a cautionary tale of the problems and pitfalls of such directions. Recently, the Victorian government has passed the Jury Directions Act 2013. This Act sets out “guiding principles” that should determine the content, and use, of jury directions. These guiding principles should form the basis for any jury directions with respect to sexual offences.

A stochastic time-domain model for burst data aggregation in IEEE 802.15.4 wireless sensor networks

In many network applications, the nature of traffic is of burst type. Often, the transient response of network to such traffics is the result of a series of interdependant events whose occurrence prediction is not a trivial task. The previous efforts in IEEE 802.15.4 networks often followed top-down approaches to model those sequences of events, i.e., through making top-view models of the whole network, they tried to track the transient response of network to burst packet arrivals. The problem with such approaches was that they were unable to give station-level views of network response and were usually complex. In this paper, we propose a non-stationary analytical model for the IEEE 802.15.4 slotted CSMA/CA medium access control (MAC) protocol under burst traffic arrival assumption and without the optional acknowledgements. We develop a station-level stochastic time-domain method from which the network-level metrics are extracted. Our bottom-up approach makes finding station-level details such as delay, collision and failure distributions possible. Moreover, network-level metrics like the average packet loss or transmission success rate can be extracted from the model. Compared to the previous models, our model is proven to be of lower memory and computational complexity order and also supports contention window sizes of greater than one. We have carried out extensive and comparative simulations to show the high accuracy of our model.

A model of pilgrimage tourism: process, interface, people and sequence

Purpose – This paper aims to examine the elements of the main process of pilgrimage tourism (PT), occurring between pilgrims, hikers and tourists along a trail towards a holy site. PT is defined as a process consisting of three sub-processes over time and across contexts: pre-process, main process and post-process. Design/methodology/approach – Explores the core reasons for PT through active participation and observation. Findings – This study reveals different layers, levels, views, approaches and perspectives involved in people-based processes. The study attempts to conceptualize the elements involved between people committed and dedicated to PT. Research limitations/implications – The introduced model of PT stresses the processes and interfaces involved over time and across contexts between people, with the same or different sequences. There is, to the best of the authors’ knowledge, no previous research that explores and describes the processes and interaction between pilgrims, hikers and tourists. Practical implications – The ultimate experience at an individual level differs, depending upon the outcome of the PT-elements of the model of PT (i.e. processes, interfaces, people and sequences). Social implications – From a social science perspective, the research examines the motives of different traveller types and looks at their different perspectives of being involved with the same physical activity of travel. The study emphasises that we can be involved in the same physical activity, but embrace it with different levels of personal and emotional engagement. Originality/value – A conceptualized model of PT containing four elements (process, interface, people and sequence) – all of which offer a foundation for structuring and assessing empirical research, and provide additional insights and knowledge into the dynamics and complexity involved specifically in a people-based process consisting of interfaces and sequences when travelling.

An improved consensus-based group decision making model with heterogeneous information

In group decision making (GDM) problems, it is natural for decision makers (DMs) to provide different preferences and evaluations owing to varying domain knowledge and cultural values. When the number of DMs is large, a higher degree of heterogeneity is expected, and it is difficult to translate heterogeneous information into one unified preference without loss of context. In this aspect, the current GDM models face two main challenges, i.e., handling the complexity pertaining to the unification of heterogeneous information from a large number of DMs, and providing optimal solutions based on unification methods. This paper presents a new consensus-based GDM model to manage heterogeneous information. In the new GDM model, an aggregation of individual priority (AIP)-based aggregation mechanism, which is able to employ flexible methods for deriving each DM's individual priority and to avoid information loss caused by unifying heterogeneous information, is utilized to aggregate the individual preferences. To reach a consensus more efficiently, different revision schemes are employed to reward/penalize the cooperative/non-cooperative DMs, respectively. The temporary collective opinion used to guide the revision process is derived by aggregating only those non-conflicting opinions at each round of revision. In order to measure the consensus in a robust manner, a position-based dissimilarity measure is developed. Compared with the existing GDM models, the proposed GDM model is more effective and flexible in processing heterogeneous information. It can be used to handle different types of information with different degrees of granularity. Six types of information are exemplified in this paper, i.e., ordinal, interval, fuzzy number, linguistic, intuitionistic fuzzy set, and real number. The results indicate that the position-based consensus measure is able to overcome possible distortions of the results in large-scale GDM problems.

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.

Time cumulative complexity modeling and analysis for space-based networks

In this paper, the notion of the cumulative time varying graph (C-TVG) is proposed to model the high dynamics and relationships between ordered static graph sequences for space-based information networks (SBINs). In order to improve the performance of management and control of the SBIN, the complexity and social properties of the SBIN's high dynamic topology during a period of time is investigated based on the proposed C-TVG. Moreover, a cumulative topology generation algorithm is designed to establish the topology evolution of the SBIN, which supports the C-TVG based complexity analysis and reduces network congestions and collisions resulting from traditional link establishment mechanisms between satellites. Simulations test the social properties of the SBIN cumulative topology generated through the proposed C-TVG algorithm. Results indicate that through the C-TVG based analysis, more complexity properties of the SBIN can be revealed than the topology analysis without time cumulation. In addition, the application of attack on the SBIN is simulated, and results indicate the validity and effectiveness of the proposed C-TVG and C-TVG based complexity analysis for the SBIN.