CONTENTS Page ACKNOWLEDGEMENT . . . . . . . . . . . . . . . . . . . . iii LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . vi CHAPTER I. INTRODUCTION . . . . . . . . . . . . . . . . 1 1.1 An Overview . . . . . . . . . . . . . . . 1 1.2 Evolution of Computer-Based Instruction at Rush . . . . . . . . . . . . . . . . . 3 1.3 Goals of the Thesis . . . . . . . . . . . 4 1.4 Organization of the Thesis . . . . . . . . 6 II. THE BACKGROUND . . . . . . . . . . . . . . . 9 2.1 Qualitative Reasoning . . . . . . . . . . 9 2.2 Subject Area . . . . . . . . . . . . . . 10 2.3 Organization . . . . . . . . . . . . . . 12 2.4 System Constraints . . . . . . . . . . . 14 2.5 Multiple Simultaneous Inputs . . . . . . . 15 III. ORGANIZATION OF CIRCSIM-TUTOR . . . . . . . . 18 3.1 Intelligent Tutoring Systems . . . . . . . 18 3.2 Domain Expertise . . . . . . . . . . . . 23 3.3 Input-Understander . . . . . . . . . . . 26 3.4 Student Modeler . . . . . . . . . . . . . 27 3.5 Instructional Planner . . . .

To operate successfully in a complex world, intelligent agents must exhibit contextsensitive behavior. Context impacts the appropriateness of virtually all aspects of an agent's behavior, yet most existing reasoning approaches pay little if any attention to explicitly recognizing, reasoning about, and making use of knowledge about the current context. We have developed a mechanism as part of our work on schema-based reasoning that uses contextual schemas (c-schemas) to explicitly represent contexts an agent may encounter. The agent's context manager retrieves the best c-schemas from its memory based on features of its current situation, then merges them to form a view of the current context, the current c-schema. This is then used to set behavioral parameters, initiate and terminate context-specific actions, focus its attention on appropriate goals to achieve, select actions for achieving them, and rapidly and appropriately handle unanticipated events. An early version of this approach...

The data exploration task can be divided into three interrelated subtasks: (i) feature selection, (ii) discovery, and (iii) interpretation. This paper describes an unsupervised discovery method with biases geared toward partitioning objects into clusters that improve interpretability. The algorithm, ITERATE, employs: (i) a data ordering scheme and (ii) an iterative redistribution operator to produce maximally cohesive and distinct clusters. Cohesion or intra-class similarity is measured in terms of the match between individual objects and their assigned cluster prototype. Distinctness or inter-class dissimilarity is measured by an average of the variance of the distribution matchbetween clusters. We demonstrate that interpretability, from a problem solving viewpoint, is addressed by theintra- and interclass measures. Empirical results demonstrate the properties of the discovery algorithm, and its applications to problem solving.

Abstract. Producing estimates of classification confidence is surprisingly difficult. One might expect that classifiers that can produce numeric classification scores (e.g. k-Nearest Neighbour or Naive Bayes) could readily produce confidence estimates based on thresholds. In fact, this proves not to be the case, probably because these are not probabilistic classifiers in the strict sense. The numeric scores coming from k-Nearest Neighbour or Naive Bayes classifiers are not well correlated with classification confidence. In this paper we describe a case-based spam filtering application that would benefit significantly from an ability to attach confidence predictions to positive classifications (i.e. messages classified as spam). We show that ‘obvious ’ confidence metrics for a case-based classifier are not effective. We propose an ensemble-like solution that aggregates a collection of confidence metrics and show that this offers an effective solution in this spam filtering domain. 1

This paper presents Bayesian networks as a framework for distributed reasoning in expert systems. We discuss methods for evidence propagation, for learning, with emphasis on sequential learning, and for generating linguistic explanations. When a parallel implementation is possible, we describe the computational power, i.e. the information that must be stored and the local calculations that must be performed at every node, in order to get a distributed expert system. Finally, a brief comparison to neural networks is offered.

This paper describes the interactive refinement of formal specifications of a structured knowledge-based system. The specification contains an explicit model of the knowledge that is to be used in problem solving and its role in problem solving. This model is used as a bias in the acquisition of domain knowledge. The model acts as bias for refinement and structures the knowledge acquisition dialogue with a user. The method has been applied in a system for learning in KADS knowledgebased systems. The system learns in two cycles. During the first cycle, the performance of the knowledge-based system is improved from experience, without modifying the initial knowledge base. During the second cycle, this knowledge base is refined interactively. 1 Introduction This research concerns the use of a model of problem solving as bias in interactive refinement of knowledge-based systems. In knowledge acquisition, knowledge elicited from various sources is used to construct or revise Knowledgebased...