Publications

Traditional system specifications always abstract from resource. However, it is unavoidable to append resource constraints in later system implementation. This paper defines a hierarchy of semantic models for programme's execution with limited resources. Each model in the hierarchy is a refinement to the one defined formerly. We also investigate homorphism and algebraic laws in these models. based on that, we develop a set of program transformations to deal with the situation that resource are in shortage.

 This document is the second of a series of three recording the output of the second phase of UNU/IIST's MultiScript project. It extends the description and the formal specification of multi-directional, multi-lingual documents presented in the UNU/IIST report 105 to cover the display and printing of such documents. The third document in the series, UNU/IIST report 113, covers the creation and editing of multi-directional, multi-lingual documents.

 Reasoning about a distributed system that exhibits a combination of probabilistic and temporal behaviour does not seem to be easy with current techniques. The reason is the interaction between probability and abstraction, made worse by remote synchronisation. In this paper the recently proposed language ptsc (for probability, time and shared-variable concurrency) is extended by constructs for interleaving and local block. Both enhance a designer's ability to modularise a design; the latter also permits a design to be compared with its more abstract specification, by concealing appropriately chosen design variables. Laws of the extended language are studied and applied in a case study consisting of a faulty register-transfer-level design.

This paper compares the accuracy of Decision Tree and Bayesian Network algorithms for predicting the academic performance of undergraduate and postgraduate students at two very different academic institutes: Can Tho University (CTU), a large national university in Viet Nam; and the Asian Institute of Technology (AIT), a small international postgraduate institute in Thailand that draws students from 86 different countries. Although the diversity of these two student populations is very different, the data-mining tools were able to achieve similar levels of accuracy for predicting student performance: 73/71% for {fail, fair, good, very good} and 94/93% for {fail, pass} at the CTU/AIT respectively. These predictions are most useful for identifying and assisting failing students at CTU (64% accurate), and for selecting Very Good students for scholarships at the AIT (82% accurate). In this analysis, the Decision Tree was consistently 3-12% more accurate than the Bayesian Network. The results of these case studies give insight into techniques for accurately predicting student performance, compare the accuracy of data mining algorithms, and demonstrate the maturity of open source tools.

We define a context­sensitive temporal prob­ ability logic for representing classes of discrete­time temporal Bayesian networks. Context constraints allow inference to be focused on only the relevant portions of the probabilistic knowledge. We provide a declarative semantics for our language. We present a Bayesian network construction algorithm whose generated networks give sound and complete answers to queries. We use related concepts in logic programming to justify our approach. We have implemented a Bayesian network construction algorithm for a subset of the theory and demonstrate it's application to the problem of evaluating the effectiveness of treatments for acute cardiac conditions.

We define a language for representing context­sensitive probabilistic knowledge. A knowledge base consists of a set of universally quantified probability sentences that include context constraints, which allow inference to be focused on only the relevant portions of the probabilistic knowledge. We provide a declarative semantics for our language. We present a query answering procedure which takes a query Q and a set of evidence E and constructs a Bayesian network to compute P (QjE). The posterior probability is then computed using any of a number of Bayesian network inference algorithms. We use the declarative semantics to prove the query procedure sound and complete. We use concepts from logic programming to justify our approach.

We present a framework for representing the prob­ abilistic effects of actions and contingent treatment plans. Our language has a well­defined declarative semantics and we have developed an implemented al­ gorithm (named BNG) that generates Bayesian net­ works (BN) to compute the posterior probabilities of queries. In this paper we address the problem of pro­ jecting a contingent treatment plan by automatically constructing a structure of interrelated BNs, which we call a BN­graph, and applying the available propaga­ tion procedures on it. To address the optimal plan generation, we base our approach on the observation that normally the target plan space has a well­defined structure. We provide a language to describe plan spaces which resembles a programming language with loops and conditionals. We briefly present the proce­ dures for finding the optimal plan(s) from such speci­ fied plan spaces.

A realistic system for planning with uncertain in­ formation in partially observable domains must be able to reason about sensing actions and to condition its further actions on the sensed infor­ mation. Among implemented planning systems, we can distinguish two approaches to contingent decision­theoretic planning. The first is char­ acterized by a highly unconstrained plan space, while the second is characterized by a constrained and inflexible specification of plan space. In this paper, we take a middle ground between these two approaches that we consider to be more prac­ tical. We permit the user to specify the structure of the space of possible plans to be considered but to do so in a flexible manner. This flexibility is obtained through the use of a modular represen tation. We separate the representation of actions from the representation of domain relations and we separate those from the representation of the plan space. Actions and domain relations are represented with schematic Bayes net fragments and plan space is represented using programming lan­ guage constructs. We present a planning system that can find optimal plans given this represen­ tation.

We describe ongoing work toward development of a decision­theoretic agent to help users choose videos based on their preferences. The DIVA (Decision Theoretic Interactive Video Advisor) system elicits user preferences using a case­based technique. Hard constraints are used to permit the user to communicate temporary deviations from his basic preferences. If the user is not happy with the system's recommendations, he can provide feedback, which is used to modify the represented preferences and generate a new set of recommendations. We describe the fundamental algorithms, the implementation, and some results from some initial experimentation.

We describe ongoing work toward development of a decision­theoretic agent to help users choose videos based on their preferences. The DIVA (Decision­ Theoretic Interactive Video Advisor) system elicits user preferences using a case­based technique. Hard constraints are used to permit the user to communi­ cate temporary deviations from his basic preferences. If the user is not happy with the system's recommen­ dations, he can provide feedback, which is used to modify the represented preferences and generate a new set of recommendations. We describe the fundamen­ tal algorithms, the implementation, and some results from some initial experimentation.

 The aim of this report is to present the basic abstractions and basic restrictions of transport domains and construct their various models. Among them we determine evolutionary transition models and general transport models which describe transport domains at the highest levels of abstraction and generality. Spatial, timed, and controlled transport models are also considered. The models constructed can be represented by associated ``transport'' logics. Models are built in a ``top-down'' style according to LSD principles. RSL and the RAISE tools are used to describe the models constructed.

Orchestration of web services using low-level languages like BPEL is akin to performing Software Engineering with just the target programming language. Orc has been proposed as an elegant and more abstract design language. But much of its elegance derives from its functional style which must be matched, by the Software Engineer, with the state-based nature of systems as a whole. This paper explores the interface between the state-based and functional views of orchestration by studying a more abstract semantics for Orc. Sound laws are established and the resulting formalism is used on examples to demonstrate the top-down development of implementations.

 This document is about the elaboration of a domain theory. It focuses on principles and methodological aspects following Michael Jackson's work. It tries to draw some guidelines on the path from informal domain to a formal domain theory on which further software development steps can be tackled and which allows consistency checking. To make the analysis practical, the observations are based on the specific domain of resources management.