Financial trading analytics increasingly demands modular, explainable, and adaptive intelligence systems capable of handling volatile market conditions and multimodal data streams. Recent advances in generative Artificial Intelligence (AI), Large Language Models (LLMs), and graph-based representations have enabled the creation of intelligent agents that can reason over complex, interconnected financial data. We introduce GWise, a graph-structured, generative AI-enabled multi-agent framework for real-time financial trading analytics delivered via secure web services. GWise models financial decision making as a directed computational graph of specialized analytical agents, including technical, fundamental, sentiment, and risk analysis crews, whose outputs are orchestrated through a memory-augmented LLM. This graph-structured design enables transparent, adaptive, and explainable trade recommendations that evolve over time. We demonstrate how the framework's agent orchestration forms a dynamic service graph, facilitating composability, fault isolation, and scalable deployment through cloud-native APIs. Extensive back-testing and simulated market conditions show that GWise outperforms traditional strategies in risk-adjusted returns while offering improved interpretability and service robustness. Our work illustrates how graph-based multiagent coordination and generative reasoning can advance real-time financial analytics as a service.

Feedback from industry on graduate Software Engineers found that students lack an academic foundation on broader systems design and development beyond the isolated nature of university assignments. To solve this, we introduced a workshop course focused on DevOps into the degree. This paper outlines the design and delivery of a course on maintaining a DevOps cloud ecosystem, simulating a real-world company environment in which students contribute software. The course proved highly successful, with areas to improve in the future surrounding the logistics of maintaining a large ecosystem, and providing technical support to students not familiar with complex toolchains.

Abstract Big data analytics technologies are rapidly expanding across all industry sectors as organisations try to make analytics an integral part of their everyday decision-making. Although there are many software tools and libraries to assist analysts and software engineers in developing solutions, organisations are looking for flexible analytics platforms that can address their specific objectives and requirements. To minimise costs, such platforms also need to co-exist with existing IT infrastructures and reuse knowledge and resources already accumulated within the organisation. To address such needs, this paper proposes the Data Analytics Solution Engineering (DASE) framework—a knowledge-driven approach supported by semantic web technologies for requirements engineering, design and development of new data analytics platforms. It includes a meta-model that captures data analytics platform requirements via a Knowledge Base, a set of guidelines that organisations can follow in engineering data analytics platforms and a reference architecture that demonstrates how to use these guidelines. We evaluate the DASE framework through two case studies and demonstrate how it can facilitate knowledge-based and requirements-driven data analytics platform engineering. The resulting data analytics platforms are observed to be user friendly, easy to maintain and flexible in handling changes to requirements. This work contributes to the body of knowledge in knowledge-driven requirements engineering, and data analytics platform engineering by providing a meta-model and a reference architecture that can be tailored to different analytics application domains.

The lack of a common framework often complicates the process of provider selection and marginal resource allocation decision. The nonlinear relationships among selection criteria greatly impact the decision-making process. The paper address the critical issue by proposing a centralised Quality of Experience (QoE) and Quality of Service (QoS)- CQoES framework. The framework considers customised priority criteria, determine the relative importance of each criterion and intelligently assign relative weights to each criterion. The framework assists the service provider to in decision making for marginal resources. To achieve the objective, we employ the Analytical Hierarchical Process (AHP), Induced OWA (IOWA) operator, Probabilistic OWA (POWA) operator, user-based collaborative filtering method with enhanced top KNN algorithm. The method handles complex nonlinear relationship of the selection criteria. It signifies consumer's customised criteria in relation to other criteria, then reorders inputs based on the ordered-inducing variable. The proposed method smartly unifies the provider's probabilistic information and the attitudinal characteristics for marginal resource allocation. To demonstrate the effectiveness of the approach, we present two scenarios and use a real cloud and other web service dataset. The experimental results show that the proposed system handles the issue of service selection and marginal resource allocation decision.

Federated learning (FL) is a promising technique for collaboratively training machine-learning models on massively distributed clients data under privacy constraints. However, the existing FL literature focuses on speeding up the learning process and ignores minimizing the communication cost which is critical for resource-constrained clients. To this end, in this article, we propose a novel 3-way hierarchical framework (THF) to promote communication efficiency in FL. Using the proposed framework, only a cluster head (CH) communicates with the cloud server through edge aggregation in order to minimize the communication cost of clients. In particular, the clients upload their local models to their respective CHs, which are responsible to forward them to the corresponding edge server. The edge server averages the local models and iterates until it achieves the edge accuracy. Afterward, each edge server uploads the edge models to the cloud server for global aggregation. In this way, model downloading and uploading requires less bandwidth due to the short distance from source to destination that makes an efficient 3-way hierarchical network structure. In addition, we formulate a joint communication and computation resource management scheme through efficient client selection in order to achieve global cost minimization in FL. We conduct extensive empirical evaluations on diverse data learning tasks on multiple data sets to signify that THF achieves global cost savings and converges within fewer communication rounds compared to other FL approaches.

Quality of Service (QoS) is the key parameter to measure the overall performance of service-oriented applications. In a myriad of web services, the QoS data has multiple highly sparse and enormous dimensions. It is a great challenge to reduce computational complexity by reducing data dimensions without losing information to predict QoS for future intervals. This paper uses an Induced Ordered Weighted Average (IOWA) layer in the prediction layer to lessen the size of a dataset and analyse the prediction accuracy of cloud QoS data. The approach enables stakeholders to manage extensive QoS data better and handle complex nonlinear predictions. The paper evaluates the cloud QoS prediction using an IOWA operator with nine neural network methods-Cascade-forward backpropagation, Elman backpropagation, Feedforward backpropagation, Generalised regression, NARX, Layer recurrent, LSTM, GRU and LSTM-GRU. The paper compares results using RMSE, MAE, and MAPE to measure prediction accuracy as a benchmark. A total of 2016 QoS data are extracted from Amazon EC2 US-West instance to predict future 96 intervals. The analysis results show that the approach significantly decreases the data size by 66%, from 2016 to 672 records with improved or equal accuracy. The case study demonstrates the approach's effectiveness while handling complexity, reducing data dimension with better prediction accuracy.

This paper proposes an advanced, open architecture to augment streaming data platforms with both complex event processing (CEP) and predictive machine learning models. We leverage the power of CEP to preprocess streams using sophisticated event pattern expressions then present these preprocessed streams for downstream training and predictive computations. We demonstrate this approach using specific technology components.

Machine learning and data analysis are becoming an essential part of the decision-making process in modern organizations. Even though new and improved analytics algorithms are developed frequently, organizations are struggling to develop analytics applications that can stay up-to-date with changing business requirements and technology innovations. The rapid development of ad-hoc programs to conduct machine learning tasks at hand has resulted in creating more expenses and efforts in the long term, a phenomenon referred to as technical debt in literature. This paper addresses the technical debt associated with data analytics applications by proposing a knowledge repository that captures analytics-related knowledge, which can be developed and maintained separately from the organization's IT infrastructure and used to design analytics applications with visual interfaces. This way, organizations can develop dynamic and adaptable analytics applications with easy-to-follow front-ends and can accommodate new data sources or machine learning models. We evaluate the proposed approach by conducting a case study that develops an application for the acquisition and management of high-frequency financial market data.

Applying machine learning techniques over streaming data is notoriously difficult as it involves the interplay of several technologies that need to be judiciously put together by IT experts. This has motivated the need to provide intuitive and easy to use interactive interfaces for financial experts to be able to leverage both Complex Event Processing (CEP) and Machine Learning (ML) technologies. The solution proposed in this paper is based on an open architecture that uses CEP engines as a pre-processing function for downstream ML training and prediction computations. We demonstrate this approach using a few scenarios involving the analysis of financial market data streams.

Enterprises today are presented with a plethora of data, tools and analytics techniques, but lack systems which help analysts to navigate these resources and identify best fitting solutions for their analytics problems. To support enterprise-level data analytics research, this paper presents Research Variable Ontology (RVO), an ontology designed to catalogue and explore essential data analytics design elements such as variables, analytics models and available data sources. RVO is specialised to support researchers with exploratory and predictive analytics problems, popularly practiced in economics and social science domains. We present the RVO design process, its schema, how it links and extends existing ontologies to provide a holistic view of analytics related knowledge and how data analysts at the enterprise level can use it. Capabilities of RVO are illustrated through a case study on House Price Prediction.

The cloud computing literature provides various ways to utilise cloud services, each with a different viewpoint and focus and mostly using heterogeneous technical-centric terms. This hinders efficient and consistent knowledge flow across the community. Little, if any, research has aimed on developing an integrated process model which captures core domain concepts and ties them together to provide an overarching view of migrating legacy systems to cloud platforms that is customisable for a given context. We adopt design science research guidelines in which we use a metamodeling approach to develop a generic process model and then evaluate and refine the model through three case studies and domain expert reviews. This research benefits academics and practitioners alike by underpinning a substrate for constructing, standardising, maintaining, and sharing bespoke cloud migration models that can be applied to given cloud adoption scenarios.

Cloud computing provides on-demand access to affordable hardware (e.g., multi-core CPUs, GPUs, disks, and networking equipment) and software (e.g., databases, application servers and data processing frameworks) platforms with features such as elasticity, pay-per-use, low upfront investment and low time to market. This has led to the proliferation of business critical applications that leverage various cloud platforms. Such applications hosted on single/multiple cloud provider platforms have diverse characteristics requiring extensive monitoring and benchmarking mechanisms to ensure run-time Quality of Service (QoS) (e.g., latency and throughput). This paper proposes, develops and validates CLAMBS-Cross-Layer Multi-Cloud Application Monitoring and Benchmarking as-a-Service for efficient QoS monitoring and benchmarking of cloud applications hosted on multi-clouds environments. The major highlight of CLAMBS is its capability of monitoring and benchmarking individual application components such as databases and web servers, distributed across cloud layers (*-aaS), spread among multiple cloud providers. We validate CLAMBS using prototype implementation and extensive experimentation and show that CLAMBS efficiently monitors and benchmarks application components on multi-cloud platforms including Amazon EC2 and Microsoft Azure.

The ability to understand application performance and pro-actively manage system state is becoming increasingly important as infrastructure services move toward commodisation models such as cloud computing. The complexity of systems being monitored in large corporations means that detailed component-by-component analysis and/or simulation of behaviour can be impractical. The key problem is to find practical systems management approaches that enable behavioural profiling and detection of anomalous events in complex real world systems. This paper details a system performance measurement method that enables slowdown event detection and characterisation of application behaviour. These measurement techniques can underpin IT service management frameworks such as the IT Infrastructure Library. This enables business to manage the stability and performance of the end-user experience in order to support productivity and effectiveness of business processes. This paper examines anomaly detection efficiency in two case studies using six detection models in a large Australian financial services organisation. The whole of service anomaly detection methodology proposed in this paper was found to be more efficient than alternative individual transaction and whole of service models.

This paper describes key modelling concepts for events, event patterns and related concepts needed to develop a distributed software framework for real-time business analytics. These concepts are specified by means of a minimal meta-model, whose implementation can enable better interoperability between different event processing systems. This in turn can support better, distributed, collaborative analytics applications in many domains. We show an implementation of our solution approach using a case study of several business analytics problems in finance.

The advancement of mobile cloud computing (MCC) has the potential to improve communication and information during disaster emergency operation. Many studies from different countries have addressed different approaches to implement mobile cloud technologies during disaster operation. However, there has been a lack of any consolidated evidence-based study to evaluate the MCC from the dual perspectives of disaster emergency situation and technology. In this paper, we provide an extensive study of mobile cloud computing research as a critical system or application based on the latest literature published from 2010 to 2015, and highlights the specific challenges in implementing mobile cloud computing in disaster emergency operation. A systematic literature review was conducted over 4 prominent journals of computer science engineering (IEEE, ACM, ProQuest, Inspec/Elsevier) and the findings were classified into four key challenges of MCC implementation during disaster emergency operation. We present a taxonomy based on the key challenges in this area, and discuss the different solutions taken to solve the challenges. We conclude the paper with a critical analysis of the challenges that have not been solved, and highlight directions for future work.

Summary One of the main activities in data‐intensive science is data analysis. Although there are many popular technologies that can assist scientists in various isolated aspects of data analysis, supporting analysis processes in holistic ways that promote system interoperability, integration and automation, as well as scientific reproducibility and efficient data handling, presents many challenges. A common solution to address these challenges is to find efficient ways of integrating various existing technologies together to meet the analysis needs of scientists (which is similar to the idea behind science gateways). We believe that this solution is essentially an exercise in software design; and in many situations, these challenges should be tackled from a software design perspective. Consequently, this paper reviews different architectural design approaches that can be used to address these challenges and proposes a service‐oriented framework called the Ad Hoc Data Grid Environment, which consists of an architectural pattern and its associated operational guidelines. The guidelines prescribe a number of activities based on an iterative decomposition approach to produce and evolve software architectures according to constantly changing user needs. The framework is demonstrated on a case study involving analysis processes required for conducting financial event studies. Copyright © 2014 John Wiley & Sons, Ltd.

The service delivery model of cloud computing acts as a key enabler for big data analytics applications enhancing productivity, efficiency and reducing costs. The ever increasing flood of data generated from smart phones and sensors such as RFID readers, traffic cams etc require innovative provisioning and QoS monitoring approaches to continuously support big data analytics. To provide essential information for effective and efficient bid data analytics application QoS monitoring, in this paper we propose and develop CLAMS-Cross-Layer Multi-Cloud Application Monitoring-as-a-Service Framework. The proposed framework: (a) performs multi-cloud monitoring, and (b) addresses the issue of cross-layer monitoring of applications. We implement and demonstrate CLAMS functions on real-world multi-cloud platforms such as Amazon and Azure.

Literature has repeatedly stressed the importance of the association between the oral health and general health of an individual and for the respective healthcare practitioners to work collaboratively. However, the absence of a knowledge base of the scientific evidence of associations between the two domains of medical and oral health has contributed to the problem of fragmented healthcare delivery. Hence, capturing and formally representing this cross-domain knowledge is vital for next generation health information systems (HIS) to become interoperable and perform decision support tasks. To this end, ontologies have been increasingly used in HIS for representing domain knowledge; but since no comprehensive ontology relevant to both the medical and oral health domains exists no knowledge base exists either. In this paper, we present our work on addressing this issue by building an Oral-Systemic Health Cross-domain Ontology (OSHCO) to capture the inter-dependent conditions across both the domains. We further extend the ontology with rules to add expressivity and derive inferences over instance data.

Abstract A Case‐Based Reasoning (CBR) system for medical diagnosis mimics the way doctors make a diagnosis. Given a new case, its accuracy in practice depends on successful retrieval of similar cases. CBR systems have had some success in dealing with simple diseases because of the robustness of their case base. However, their diagnostic accuracy suffers when dealing with complex diseases particularly those that involve multiple domains in medicine. An example of such a condition is Premenstrual syndrome (PMS) as it falls under both gynaecology and psychiatry. To address this issue, the paper proposes a CBR‐based expert system that uses the K‐nearest neighbour (KNN) algorithm to search k similar cases based on the Euclidean distance measure. The novelty of the system is in the design of a flexible auto‐set tolerance (T), which serves as a threshold to extract cases for which similarities are greater than the assigned value of T. A prototype software tool with a menu‐driven Graphical User Interface (GUI) has been developed for case input, analysis of results, and case adaptation within the system. Finally, the performance of the tool has been checked on a set of real‐world PMS cases.

Cloud monitoring involves dynamically tracking the Quality of Service (QoS) parameters related to virtualized services (e.g., CPU, storage, network, appliances, etc.), the physical resources they share, and the applications running on them or data hosted on them. Monitoring techniques and services can help a cloud provider or application developer in regards to: (i) keeping the cloud services and hosted applications operating at peak efficiency; (ii) detecting variations in service and application performance; (iii) accounting the SLA violations of certain QoS parameters; and (iv) tracking the leave and join operations of cloud services due to failures and other dynamic configuration changes. In this paper, we describe the PhD research motivation, question, and approach and methodology related to developing novel cloud monitoring techniques and services enabling automated application QoS management under uncertainties.

The paradigm of service-oriented computing (SOC) has emerged as an architectural approach to flexibility and agility, not just in systems development but also in business process management. There is, however, a paucity of critical research assessing the strategic impact of SOA on the competitiveness of organizations. Some research literature in strategic management indicates that firms may gain a competitive advantage in rapidly changing market environments by concentrating on their dynamic capabilities - i.e., product flexibility and agility in organizational transformation in response to rapidly changing market conditions and customer requirements. The intent of this paper is to analyze the conduits through which service-oriented architectures (SOAs) may exert influence on dynamic capabilities within firms. The results could potentially assist in evaluating if and how the adoption of service-oriented architecture may help achieve key dynamic capabilities, giving the enterprise a competitive edge.

Architecting distributed software applications is a complex design activity. It involves making decisions about a number of inter-dependent design choices that relate to a range of design concerns. Each decision requires selecting among a number of alternatives; each of which impacts differently on various quality attributes. Additionally, there are usually a number of stakeholders participating in the decision-making process with different, often conflicting, quality goals, and project constraints, such as cost and schedule. To facilitate the architectural design process, we propose a quantitative quality-driven approach that attempts to find the best possible fit between conflicting stakeholders' quality goals, competing architectural concerns, and project constraints. The approach uses optimization techniques to recommend the optimal candidate architecture. Applicability of the proposed approach is assessed using a real system.

Internet-based technologies have opened new opportunities for conducting business within and across enterprises that were never possible a few years ago. This paper presents our experience in using Web services for prototyping a service-oriented architecture for capital market systems (CMSs). Our work exposes a world-class surveillance system's functionality into a number of Web services. Our work also includes benchmarking the performance of this legacy system and investigating the associated overheads of using SOAP as a wire format for Web services. Even though other research studies have tried to explain SOAP's performance inefficiency, there is lack of studies that evaluate SOAP in the context of a realistic business application. This initial investigation shows that system's integration opportunities introduced by Web services can outweigh the performance overheads. This occurs in some aspects of real-time CMSs that are not performance-demanding such as the dissemination of market alerts to the analysts.

As use of object orientation for application development has increased, many researchers have investigated the design of object-based programming languages for distributed and persistent programming. This paper concentrates on reviewing a number of such languages for distributed or persistent programming. In each case, the focus is on the object model supported and the mechanisms and policies employed in the implementation of distributed or persistent objects. In particular, each language reviewed has been chosen to illustrate a particular object model or implementation strategy.

Abstract This paper is concerned with the implementation of functional languages on a parallel architecture, using graph reduction as a model of computation. Parallelism in such systems is automatically derived by the compiler but a major problem is the fine granularity, illustrated in Divide‐and‐Conquer problems at the leaves of the computational tree. The paper addresses this issue and proposes a method based on static analysis combined with run‐time tests to remove the excess in parallelism. We report experiments on a prototype machine, simulated on several connected INMOS transputers. Performance figures show the benefits in adopting the method and the difficulty of automatically deriving the optimum partitioning due to differences among the problems.