Context Modeling

The expanding availability of high bandwidth networking is heralding a new generation of resources targeting very tightly-coupled distributed interactions. A case in point is the virtualization of critical tasks: advanced training, hazardous industrial operations, or minimally invasive surgery all require that complex entities -mechanical, thermal, physiological, etc. -be blended into cohesive "worlds" and made available to multiple actors requiring different levels of representation. In this paper we discuss issues related to the integration of computer vision and behavioral modeling, hard real-time compliance, the architecture of distributed simulation and an object-oriented framework with an extensible knowledge of physics, leading to the design of Extreme Virtual Environments. We also present current exemples targeting underwater robotics and computer-aided surgery.

a set of samples which have small diflerences due to noisy pixels. The learned models can be used to identify objects in outdoors i m g e s irrespectively of their size and partial occlusions. Some results of the inference procedure are shown in the pagel:

Pictorial structure models are the de facto standard for 2D human pose estimation. Numerous refinements and improvements have been proposed such as discriminatively trained body part detectors, flexible body models, and local and global mixtures. While these techniques allow to achieve state-of-the-art performance for 2D pose estimation, they have not yet been extended to enable pose estimation in 3D. This paper thus proposes a multi-view pictorial structures model that builds on recent advances in 2D pose esti-mation and incorporates evidence across multiple viewpoints to allow for robust 3D pose estimation. We evaluate our multi-view pictorial structures approach on the HumanEva-I and MPII Cooking dataset. In comparison to related work for 3D pose estimation our approach achieves similar or better results while operating on single-frames only and not relying on activity specific motion models or tracking. Notably, our approach outper-forms state-of-the-art for activities with more...

In many fields, such as language acquisition, neuropsychology of language, the study of ageing, and historical linguistics, corpora are used for estimating the diversity of grammatical structures that are produced during a period by an individual, community, or type of speakers. In these cases, treebanks are taken as representative samples of the syntactic structures that might be encountered. Generalizing the potential syntactic diversity from the structures documented in a small corpus requires careful extrapolation whose accuracy is constrained by the limited size of representative sub-corpora. In this article, I demonstrate—both theoretically and empirically—that a grammar’s derivational entropy and the mean length of the utterances (MLU) it generates are fundamentally linked, giving rise to a new measure, the derivational entropy rate. The mean length of utterances becomes the most practical index of syntactic complexity; I demonstrate that MLU is not a mere proxy, but a fundamental measure of syntactic diversity. In combination with the new derivational entropy rate measure, it provides a theory-free assessment of grammatical complexity. The derivational entropy rate indexes the rate at which different grammatical annotation frameworks determine the grammatical complexity of treebanks. I evaluate the Smoothed Induced Treebank Entropy (SITE) as a tool for estimating these measures accurately, even from very small treebanks. I conclude by discussing important implications of these results for both NLP and human language processing.

Journal Article by Dr Fermín Moscoso del Prado Martín published in Computational Linguistics, presented by Suchir Salhan at ACL 2025, Vienna Austria.

While focusing on a holistic and integrated view of an organization, enterprise architectures can be subdivided into components such as business architecture, data architecture, application architecture, and technology architecture in order to contain the complexity of the design problem. This wide scope of systems, requires a great flexibility in technology and management that is not necessarily required in systems with a more narrow impact. This minitrack seeks to investigate issues specific to architecture, implementation, and infrastructure for enterprisewide systems. The diversity of topics represented in the three papers in this area are testament to the breadth of areas that impact upon enterprise systems. Armour, Kaisler, Getter, and Pippin in their paper titled: "A UML-driven Enterprise Architecture Case Study" address the issues of UML modeling of US Capitol Police enterprise management system modeling. The authors developed a comprehensive set of context, process and use cases models for the overall USCP enterprise as well as each of the USCP administrative BAPs and several of the law enforcement systems. These models were used to validate the requirements, interrelationships, and constraints of each information system with the stakeholders. The authors created an enterprise system architecture that is driving the remediation, renovation, and replacement of USCP information systems. The future research directions include integration of agent markup languages (specifically DAML) for modeling interaction content and investigation of better mechanism for modeling the dynamic properties of business systems through further extensions to UML. Sridhar and Dutta look at the issues of technology proliferation in developing countries in "Modeling The Growth of Cellular Services in India: A System Dynamics Approach." Models for technology proliferation that apply in the United states and other more developed countries are not necessarily applicable to other locales which may have a monopoly or oligopoly structure. The authors use the systems dynamics methodology to develop an initial model of cellular service provision within the duopoly market structure. Because the basic construct of a systems dynamics model is the feedback loop, insights into causal mechanisms of growth are obtained as a byproduct of model structure. They describe preliminary tests of the model using subscriber growth data for the Indian cellular industry, perform sensitivity analyses to demonstrate the model's potential as a decision support tool, and conclude by identifying model extensions that would help capture this and similar evolving oligopoly market structures. The underlying networks that support larger enterprises may become quite complex and require a specific focus. Kawatra takes a mathematical approach in "A Hop Constrained MIN-SUM Arborescence with Outage Costs." He analyzes the arborescence problem that is frequently encountered in the network design, routing and scheduling. His proposal consists of finding links to connect a set of geographically remote nodes to a central node such that for each remote node j there is exactly one link entering node j, and for each remote node, a unique path exists from the central node to node j. He studies the problem of selecting links in a min-sum arborescence with hop constraints problem where each terminal node has an associated outage cost. The location of the central node and terminal nodes, and the limit h on the number of hops between the central node and each terminal node is given. The problem is formulated as an integer programming problem and a Lagrangian-based heuristic method is applied to solve it. The quality of this heuristic solution is then estimated using the lower bound given by the Lagrangian method. Subgradient optimization method is further used to find good lower bounds.

The complexity of software in embedded systems has increased significantly over the last years so that software verification now plays an important role in ensuring the overall product quality. In this context, SAT-based bounded model checking has been successfully applied to discover subtle errors, but for larger applications, it often suffers from the state space explosion problem. This paper describes a new approach called continuous verification to detect design errors as quickly as possible by looking at the Software Configuration Management (SCM) system and by combining dynamic and static verification to reduce the state space to be explored. We also give a set of encodings that provide accurate support for program verification and use different background theories in order to improve scalability and precision in a completely automatic way. A case study from the telecommunications domain shows that the proposed approach improves the error-detection capability and reduces the overall verification time by up to 50%.

Piece-wise (PW) behavioral models are commonly adopted when modeling RF power amplifiers (PAs) that exhibit strong amplitude dependent nonlinear distortion characteristics. In this paper, we consider a new PW model for PAs based on the mixture of experts (ME) approach. We first introduce the ME framework theory while also extend it such that it can be applied to model complex baseband signals and nonlinearities. Then, we show how the ME model allows overcoming some of the intrinsic shortcomings that existing PW behavioral models commonly exhibit, which translates into improved modeling accuracy. The proposed solution is validated with extensive RF measurements on a load modulated balanced PA at 2.1 GHz, excited by a 320 MHz OFDM signal, and is benchmarked against several state-of-the art PW models.

We introduce SRI's new speech recognition engine, DynaSpeak TM , which is characterized by its scalability and flexibility, high recognition accuracy, memory and speed efficiency, adaptation capability, efficient grammar optimization, support for natural language parsing functionality, and operation based on integer arithmetic. These features are designed to address the needs of the fast-developing and changing domain of embedded and mobile computing platforms.

While context-free grammars (CFGs) remain as one of the most important formalisms for interpreting natural language, word ngram models are surprisingly powerful for domain-independent applications. We propose to unify these two formalisms for both speech recognition and spoken language understanding (SLU). With portability as the major problem, we incorporated domainspecific CFGs into a domain-independent n-gram model that can improve generalizability of the CFG and specificity of the n-gram. In our experiments, the unified model can significantly reduce the test set perplexity from 378 to 90 in comparison with a domainindependent word trigram. The unified model converges well when the domain-specific data becomes available. The perplexity can be further reduced from 90 to 65 with a limited amount of domain-specific data. While we have demonstrated excellent portability, the full potential of our approach lies in its unified recognition and understanding that we are investigating.

Stateflow is a visual tool that is used extensively in industry for designing the reactive behaviour of embedded systems. Stateflow relies on techniques like simulation to aid the user in finding flaws in the model. However, simulation is inadequate as a means of detecting inconsistencies and incompleteness in the model. Tabular Expressions (function tables) have been used successfully in software development for more than thirty years. Tabular expressions are also visual representations of functions, but include the important properties of completeness and disjointness. In other words, a tabular expression is well-formed only when the input domain is covered completely (completeness), and when there is no ambiguity in the behaviour described by the tabular expression (disjointness). The goal of our work is to use the completeness and disjointness properties of well-formed tabular expressions to aid us in establishing those properties in Stateflow models. From the Stateflow models, we generate a new kind of tabular expression that includes extended output options. We use the informal Stateflow semantics from MathWorks documentation as the basis for generating our tabular expressions. The generated tabular expressions are then used to guarantee completeness and disjointness. We provide a transformation algorithm that we plan to implement in a tool to automatically generate tabular expressions from Stateflow models. •Software and its engineering → Consistency; Completeness; Formal software verification; Software notations and tools; Requirements analysis; Software design engineering;

This paper deals with the problem of fault detection and identification in noisy systems. A proportionnal integral observer with unknown inputs is used to reconstruct state and sensors faults. A mathematical transformation is made to conceive an augmented system, in which the initial sensor fault appear as an unknown input. The noise effect on the state and fault estimation errors is also minimized. The obtained results are then extended to nonlinear systems described by nonlinear Takagi-Sugeno models.

This paper deals with the problem of fault detection and identification in noisy systems. A proportionnal integral observer with unknown inputs is used in order to estimate the state and the faults which are assumed as unknown inputs. The noise effect on the state and fault estimation errors is also minimized. The obtained results are then extended to nonlinear systems described by nonlinear Takagi-Sugeno models.

The focus of this paper is to describe the development and the architecture of a new parallel algorithm targeted for image compression within High Performance Computing context. The suggested algorithm apply parallel processing strategies for the image data set in order to minimize the amount of computation while at the same time optimize the compression ratio and speed. When combining new parallelism strategies with newly developed and existing methods of data compression, the result is visibly better in both, compression ratio and time, as opposed to comparable existing algorithms. Experimental results performed on different HPC environment prove that the solution put forward is quite scalable and efficient; therefore, it should be considered in applications where real time image processing is crucial. The novelty of this work lies in the design and implementation of a parallel algorithm that simultaneously integrates bitplane coding, wavelet transforms, and predictive coding within a scalable HPC framework-an approach not yet addressed comprehensively in existing methods. This hybrid strategy provides significant improvements in compression ratio, execution time, and scalability when compared to state-of-the-art image compression techniques.

We are interested in systems which do not prescribe one single kind of preference, but in which varying kinds of preferences can be used simultaneously. In such systems it is essential to know the interaction among the kinds of preferences being used, and we therefore introduce and study a nonmonotonic logic to reason about sixteen strict and non-strict kinds of preferences, including ceteris paribus preferences. Moreover, we study "distinguished" preference orders based on specificity principles by showing when these distinguished preference orders are unique, and by presenting algorithms to calculate the distinguished preference orders.

Driver's body posture in 3D contains information potentially related to driver intent, driver affective state, and driver distraction. In this paper, we discuss issues and possibilities in developing a vision-based, markerless system to systematically explore the role of 3D driver posture dynamics for driver assistance. At high level, two main emphases in the proposed system are: (i) The coordination between real world driving testbed and simulation environment and (ii) The usefulness of driver posture dynamics is studied not only as an individual cue but also in relation with other contextual information (e.g. head dynamics, facial features, and vehicle dynamics). Some initial results in our experiment following these guidelines show the feasibility and promise of extracting and using 3D driver posture dynamics for driver assistance.

We study key issues related to multilingual acoustic modeling for automatic speech recognition (ASR) through a series of large-scale ASR experiments. Our study explores shared structures embedded in a large collection of speech data spanning over a number of spoken languages in order to establish a common set of universal phone models that can be used for large vocabulary ASR of all the languages seen or unseen during training. Language-universal and language-adaptive models are compared with language-specific models, and the comparison results show that in many cases it is possible to build general-purpose language-universal and language-adaptive acoustic models that outperform language-specific ones if the set of shared units, the structure of shared states, and the shared acoustic-phonetic properties among different languages can be properly utilized. Specifically, our results demonstrate that when the context coverage is poor in language-specific training, we can use one tenth of the adaptation data to achieve equivalent performance in cross-lingual speech recognition.

No está permitida la reproducción total o parcial de este libro, ni su incorporación a un sistema informático, ni su transmisión en cualquier forma o por cualquier medio, sea este electrónico, mecánico, por fotocopia, por grabación u otros métodos, sin el permiso previo y por escrito del editor. La infracción de los derechos mencionados puede ser constitutiva de delito contra la propiedad intelectual (art. 270 y siguientes del Código Penal). Diríjase a Cedro (Centro Español de Derechos Reprográficos) si necesita fotocopiar o escanear algún fragmento de esta obra. Puede contactar con Cedro a través de la web www.conlicencia.com o por teléfono en el 917021970/932720407. Este libro ha sido sometido a evaluación por parte de nuestro Consejo Editorial. Para mayor información, véase www.dykinson.com/quienes_somos

One of the challenges in Video Event Retrieval, the field in which (a sequence of frames with) high-level events are retrieved from a set of videos, is to model the temporal structure. One way to incorporate this information is using ANDOR graphs, which is a type of graphical model consisting of layers with AND nodes and OR nodes. We introduce new nodes, such as the BEFORE and WHILE node, for AND-OR graphs to explicitly model temporal information. The advantage of these nodes is that the graph is insightful and transparent for a user. Additionally, the graph can both be created by a user or with the use of training examples. We perform initial experiments on a video surveillance dataset named VIRAT, which contains temporally inverse events with the same concepts, such as entering and exiting a building. We compare performance to state of the art Support Vector Machine and Hidden Markov Model methods. We show that our proposed graph with WHILE and BEFORE nodes outperforms the state o...

A very promising approach to increase productivity, quality and competitiveness approach of information systems development is the reuse and development of a family of systems guided by Product Line (PL) practices. One of the main goals of PL engineering is to develop a model that represents the family of products (product line model PLM), which is then customized to configure individual products. The successful definition of PLMs that accurately represents the information in the requirement specifications still depends heavily on the intuition and experience of the software architect. Our work provides assistance for this process. We have developed a semiautomated method to construct product line models based on collection of related artifacts or existent products models as a result of a feature mining process. The approach is evaluated using bill of material as a collection of product models to develop and construct a constraint based PLM. The performance of our method is calculated by estimating the time complexity and constructing the PLM for different random samples of 536 products in Baxter Bioscience. More than 92% of the relationships are properly predicted only by using 75% of the total available products.

Learning objects repositories intend to endorse reutilization but this purpose has not been fully achieved yet. In this paper, we describe a project addresses the effective reuse of learning objects in repositories. The initial model presented and discussed introduces new requirements in the way contents are stored and searched and in the contextual definition of learning objects. Considering IMS LD templates, we propose examples of its use for learning objects.

An explicit connection between fitting exponential models and pole-zero models to observed data is made. The fitting problem is formulated as a constrained nonlinear minimization problem. This problem is then solved using a simplified iterative algorithm. The algorithm is applied to simulated data, and the performance of the algorithm is compared to previous results.

Although IEEE 802.11 provides several transmission rates, a suitable rate adaptation taking into account the relative fairness among all competitive stations, according to the underlying channel quality remains a challenge in Mobile Ad hoc Networks (MANETs). The absence of any fixed infrastructure and any centralized control makes the existing solutions for WLANs like CARA (collision-aware rate adaptation) not appropriate for MANETs. In this paper, we propose a new analytical model with a suitable approach to ensure a relative fairness among all competitive nodes of a particular channel. Our model deals with the channel quality while respecting the nodes, based on transmission successes and failures in a mobility context. Finally, each node calculates its own probability to access the channel in a distributed manner. We evaluate the performance of our scheme with others in the context of MANET via extensive and detailed simulations. The performance differentials are analysed using varying network load and transmission range. The simulation results illustrate that our proposed approach ensures a better tradeoff between fairness and throughput.

In support of the AFOSR program in Information Fusion, the CNS Technology Laboratory at Boston University is developing and applying neural models of image and signal processing, pattern learning and recognition, associative learning dynamics, and 3D visualization, to the domain of Information Fusion for Image Analysis in a geospatial context. Our research is focused by a challenge problem involving the emergence of a crisis in an urban environment, brought on by a terrorist attack or other man-made or natural disaster. We aim to develop methods aiding preparation and monitoring of the battlespace, deriving context from multiple sources of imagery (high-resolution visible and low-resolution hyperspectral) and signals (GMTI from moving vehicles, and ELINT from emitters). This context will serve as a foundation, in conjunction with existing knowledge nets, for exploring neural methods in higherlevel information fusion supporting situation assessment and creation of a common operating picture (COP).

Binary images can be compressed efficiently using context-based statistical modeling and arithmetic coding. However, this approach is fully sequential and therefore additional computing power from parallel computers cannot be utilized. We attack this problem and show how to implement the context-based compression in parallel. Our approach is to segment the image into non-overlapping blocks, which are compressed independently by the processors. We give two alternative solutions about how to construct, distribute and utilize the model in parallel, and study the effect on the compression performance and execution time. We show by experiments that the proposed approach achieves speedup that is proportional to the number of processors. The work efficiency exceeds 50% with any reasonable number of processors.

A method for compressing large binary images is proposed for applications where spatial access to the image is required. The proposed method is a two-stage combination of forward-adaptive modeling and backwardadaptive context based compression with re-initialization of statistics. The method improves compression performance significantly in comparison to a straightforward combination of JBIG and tiling. Only minor modifications to the QM-coder are required, and therefore existing software implementations can be easily utilized. Technical details of the modifications are provided.

A method for lossless compression of large binary images is proposed for applications where spatial access to the image is needed. The method utilizes the advantages of (1) variable-size context modeling in a form of context trees, and (2) forward-adaptive statistical compression. New strategies for constructing the context tree are introduced, including a fast two-stage bottom-up approach. The proposed technique achieves higher compression rates and allows dense tiling of images down to 50 × 50 pixels without sacrificing the compression performance. It enables partial decompression of large images far more efficiently than if the standard JBIG was applied.

Binary images can be compressed efficiently using context-based statistical modeling and arithmetic coding. However, this approach is fully sequential and therefore additional computing power from parallel computers cannot be utilized. We attack this problem and show how to implement the context-based compression in parallel. Our approach is to segment the image into non-overlapping blocks, which are compressed independently by the processors. We give two alternative solutions about how to construct, distribute and utilize the model in parallel, and study the effect on the compression performance and execution time. We show by experiments that the proposed approach achieves speedup that is proportional to the number of processors. The work efficiency exceeds 50% with any reasonable number of processors.

A method for compressing large binary images is proposed for applications where spatial access to the image is required. The proposed method is a two-stage combination of forward-adaptive modeling and backwardadaptive context based compression with re-initialization of statistics. The method improves compression performance significantly in comparison to a straightforward combination of JBIG and tiling. Only minor modifications to the QM-coder are required, and therefore existing software implementations can be easily utilized. Technical details of the modifications are provided.

A method for lossless compression of large binary images is proposed for applications where spatial access to the image is needed. The method utilizes the advantages of (1) variable-size context modeling in a form of context trees, and (2) forward-adaptive statistical compression. New strategies for constructing the context tree are introduced, including a fast two-stage bottom-up approach. The proposed technique achieves higher compression rates and allows dense tiling of images down to 50 × 50 pixels without sacrificing the compression performance. It enables partial decompression of large images far more efficiently than if the standard JBIG was applied.

We propose a two-stage system for spoken language machine translation. In the first stage, the source sentence is parsed and paraphrased into an intermediate language which retains the words in the source language but follows the word order of the target language as much as feasible. This stage is mostly linguistic. In the second stage, a statistical MT is performed to translate the intermediate language into the target language. For the task of English-to-Mandarin translation, we achieved a 2.5 increase in BLEU score and a 45% decrease in GIZA-Alignment Crossover, on IWSLT-06 data. In a human evaluation of the sentences that differed, the two-stage system was preferred three times as often as the baseline.

Recent progress and prospects in cognitive vision, multimedia, human-computer interaction, communications and the Web call for, and can profit from applications of advanced image and video analysis. Adaptive robust systems are required for analysis, indexing and summarization of large amounts of audio-visual data. Grace aux nouvelles capacités d'apprentissage de HMM, nous croyons que ce type de modèle sera appelé à être utilisé pour nombre d'applications. On peut envisager d'autres applications 3-D telles que la classification d'images 3-D ou la reconstruction d'image. Au delà, puisque l'arbre de dépendance présente des discontinuités nous pouvons trouver d'autres façons de choisir un arbre. Faire par exemple un arbre optimal pour un jeu d'images ou application en analysant d'abord chaque image et faire plus de rapports (connexions) entre des régions de frontière, ou considérer d'autres arbres non-aléatoires. Une future approche du problème de discontinuités est d'utiliser des modèles hiérarchiques, puisqu'un bloc dans une résolution plus basse peut inclure les rapports (connexions) qui n'existent pas à plus petite échelle. Le modèle peut aussi être appliqué (étendu) aux dimensions supérieures (n > 3). Dans ce cas les rapports contextuels deviennent alors plus faibles et pour chaque dimension nous aurons un rapport (connexion)proche de n. Pour cette raison le choix de l'arbre deviendra encore plus important. En conclusion nous croyons que le DT HMM est un modèle puissant qui a un véritable potentiel d'avenir pour de nombreuses applications. This thesis is the result of three and a half years effort spent at the Multimedia Communications Department of institute Eurecom, which has provided me with enriching industrial and academic experiences. First of all, I would like to thank my principal advisor, Professor Bernard Merialdo, for his great guidance, support and patience. The weekly meetings have greatly helped in developing my work. He has provided me not only with technical and mathematical knowledge but also a rigorous attitude towards research, for which I am grateful. I also express my gratitude to assistant Professor Benoit Huet, for having many discussions with me in machine learning and image processing.

The proposed framework lets developers model and manage process variability by composing base models, fragments, and variability models and by deferring binding to run time. Base models and fragments are reusable, thereby reducing the modeling effort for developing variants. //

This thesis describes a research work aimed at improving the interfaces of online discussion forums (ODFs) in relation to their functional support to enhance learning. These ODFs form part of almost all Learning Management Systems (LMSs) such as WebCT, Moodle and Blackboard, which are widely used in education nowadays. Although ODFs are identified as valuable sources to learning, their interfaces are limited in terms of providing support to students, such as in the areas of managing their postings as well as in facilitating them to quickly locate and obtain specified information. In addition, these systems lack features to support inter-institutional cooperation that could potentially increase knowledge sharing between students and educators of different institutions. The interface design objective of this study therefore was to explore and overcome the limitations identified as above, and enhance the effectiveness and efficiency of ODFs' support to learning. Using a task centered design approach; the required features were developed, and implemented in a working prototype called eQuake (electronic Question answer knowledge environment). eQuake is a shared online discussion forum system developed as an add-on to a well-known open source e-learning platform (Moodle). This system was intended for use among interinstitutional students in New Zealand tertiary institutions that teach similar courses. The improved interface functionalities of eQuake are expected to enhance learning support in terms of widening communication among users, increasing knowledge base, providing existing matching answer(s) quickly to students, and exposing students to multiple perspectives. This study considers such improvements to ODF interfaces as vital to enable users to enjoy the benefits of technology-mediated environment. The perceived usefulness and ease-of-use of improved features in eQuake were evaluated using a quantitative experimental research method. The evaluation was conducted at three tertiary institutions in New Zealand, and the overall results indicated positive response, although some suggestions for improvement have been made in the evaluation. This thesis presents a review of the related literature, describes the design and development of a user interface, followed by its implementation in eQuake, and a description of the evaluation. The thesis concludes with recommendations for better interface design of ODFs and provides suggestions for future research in this area.

This thesis describes a research work aimed at improving the interfaces of online discussion forums (ODFs) in relation to their functional support to enhance learning. These ODFs form part of almost all Learning Management Systems (LMSs) such as WebCT, Moodle and Blackboard, which are widely used in education nowadays. Although ODFs are identified as valuable sources to learning, their interfaces are limited in terms of providing support to students, such as in the areas of managing their postings as well as in facilitating them to quickly locate and obtain specified information. In addition, these systems lack features to support inter-institutional cooperation that could potentially increase knowledge sharing between students and educators of different institutions. The interface design objective of this study therefore was to explore and overcome the limitations identified as above, and enhance the effectiveness and efficiency of ODFs' support to learning. Using a task centered design approach; the required features were developed, and implemented in a working prototype called eQuake (electronic Question answer knowledge environment). eQuake is a shared online discussion forum system developed as an add-on to a well-known open source e-learning platform (Moodle). This system was intended for use among interinstitutional students in New Zealand tertiary institutions that teach similar courses. The improved interface functionalities of eQuake are expected to enhance learning support in terms of widening communication among users, increasing knowledge base, providing existing matching answer(s) quickly to students, and exposing students to multiple perspectives. This study considers such improvements to ODF interfaces as vital to enable users to enjoy the benefits of technology-mediated environment. The perceived usefulness and ease-of-use of improved features in eQuake were evaluated using a quantitative experimental research method. The evaluation was conducted at three tertiary institutions in New Zealand, and the overall results indicated positive response, although some suggestions for improvement have been made in the evaluation. This thesis presents a review of the related literature, describes the design and development of a user interface, followed by its implementation in eQuake, and a description of the evaluation. The thesis concludes with recommendations for better interface design of ODFs and provides suggestions for future research in this area.

Context-aware Web services are emerging as a promising technology for the electronic businesses in mobile and pervasive environments. Unfortunately, complex contextaware services are still hard to build. In this paper, we present a modeling language for the model-driven development of context-aware Web services based on the Unified Modeling Language (UML). Specifically, we show how UML can be used to specify information related to the design of context-aware services. We present the abstract syntax and notation of the language and illustrate its usage using an example service. Our language offers significant design flexibility that considerably simplifies the development of context-aware Web services.

As modern information systems become increasingly business-and safety-critical, it is extremely important to improve both the trust that a user places in a system and their understanding of the risks associated with making a decision. This paper presents the STRAPP framework, a generic framework that supports both of these goals through the use of personalised provenance reasoning engines and state-of-art risk assessment techniques. We present the high-level architecture of the framework, and describe the process of systematically modelling system provenance with the W3C PROV provenance data model. We discuss the business drivers behind the concept of personalizing provenance information, and describe the STRAPP approach to enabling this through a user-adaptive system style. We discuss using data provenance for risk management and treatment in order to evaluate risk levels, and discuss the use of CORAS to develop a risk reasoning engine representing core classes and relationships. Finally, we demonstrate the initial implementation of our personalised provenance system in the context of the Rolls-Royce Equipment Health Management, and discuss its operation, the lessons we have learnt through our research and implementation (both technical and in business), and our future plans for this project.

Middleware support for pervasive context-aware systems relieves context-aware applications from dealing with the complexity of context-specific operations such as context acquisition, aggregation, reasoning and distribution. The middleware decouples applications from the underlying heterogeneous context sensors, and offers advantages such as rapid development of context-aware applications and efficient usage of the context sensors. Context sensors have inherent limitations with respect to the quality of the context information they produce. Without breaking the decoupling, the middleware needs to explicitly model and quantify this Quality of Context in order to allow application to adapt their behavior based on the Quality of Context, for efficiency reasons and to enable Quality-of-Context-aware privacy policies. In this paper we identify and define five Quality-of-Context indicators for context-aware middleware, and discuss different alternatives for their quantification. These Quality-of-Context indicators are: precision, freshness, spatial resolution, temporal resolution and probability of correctness.

In the last years, authoring based on e-learning standards has been consolidated as a core factor of industry and development of interoperable and effective virtual learning environments. However, there is a need for further research on abstraction to provide a more instructional view in the context of authoring tools in a variety of ways, in order to avoid being driven by Learning Technology (LT) specifications, facilitate instructional knowledge aggregation, and to provide an appropriate level of clarity and semantics in the design of collaborative activities. We propose a combination of techniques to provide this instructional abstraction in the context of the new European educational model, combining instructional layers and collaborative scripts in authoring tools, and semantic web techniques for extending e-learning material in order to harness the wealth of existing web content and semantically labeled repositories.

Distributed simulation has proved to be a cost effecti ve technique for studying and understanding complex realworld systems. Many distributed simulations need to incorporate hundreds or even thousands of processors, across both local and wide area networks. As the size and complexity of distributed simulations increase so do the demands on the hardware and software resources that provide simulation services, causing performance bottlenecks that limit the effectiveness of distributed simulation. Different abstractions of performance are needed depending on who is analyzing the distributed simulation and for what reason. This paper proposes a framework for identifying the factors affecting performance and provides a technique for associating the performance factors with high-level system metrics that describe the behavior of the physical and logical resources and services used in the design and implementation of distributed simulations. Dynamic and static analysis of the performance information provides feedback on the execution of the simulation and can provide meaningful information as a guide in making decisions about the configuration and control of the available hardware and software resources for distributed simulation exercises.