Case-based reasoning solves new problems by re-using the solutions of previously solved similar problems and is popular because many of the knowledge engineering demands of conventional knowledge-based systems are removed. The content of the case knowledge container is critical to the performance of case-based classification systems. However, the knowledge engineer is given little support in the selection of suitable techniques to maintain and monitor the case base. This research investigates the coverage, competence and problem-solving capacity of case knowledge with the aim of developing techniques to model and maintain the case base. We present a novel technique that creates a model of the case base by measuring the uncertainty in local areas of the problem space based on the local mix of solutions present. The model provides an insight into the structure of a case base by means of a complexity profile that can assist maintenance decision-making and provide a benchmark to assess future changes to the case base. The distribution of cases in the case base is critical to the performance of a case-based reasoning system. We argue that classification boundaries represent important regions of the problem space and develop two complexity-guided algorithms which use boundary identification techniques to actively discover cases close to boundaries. We introduce a complexity-guided redundancy reduction algorithm which uses a case complexity threshold to retain cases close to boundaries and delete cases that form single class clusters. The algorithm offers control over the balance between maintaining competence and reducing case base size. The performance of a case-based reasoning system relies on the integrity of its case base but in real life applications the available data invariably contains erroneous, noisy cases. Automated removal of these noisy cases can improve system accuracy. In addition, error rates can often be reduced by removing cases to give smoother decision boundaries between classes. We show that the optimal level of boundary smoothing is domain dependent and, therefore, our approach to error reduction reacts to the characteristics of the domain by setting an appropriate level of smoothing. We introduce a novel algorithm which identifies and removes both noisy and boundary cases with the aid of a local distance ratio. A prototype interface has been developed that shows how the modelling and maintenance approaches can be used in practice in an interactive manner. The interface allows the knowledge engineer to make informed maintenance choices without the need for extensive evaluation effort while, at the same time, retaining control over the process. One of the strengths of our approach is in applying a consistent, integrated method to case base maintenance to provide a transparent process that gives a degree of explanation.

Case-Based Reasoning (CBR) is a technique in Artificial Intelligence where a new problem is solved by making use of the solution to a similar past problem situation. People naturally solve problems in this way, without even thinking about it. For example, an occupational therapist (OT) that assesses the needs of a new disabled person may be reminded of a previous person in terms of their disabilities. He may or may not decide to recommend the same devices based on the outcome of an earlier (disabled) person. Case-based reasoning makes use of a collection of past problem-solving experiences thus enabling users to exploit the information of others’ successes and failures to solve their own problem(s). This project has developed a CBR tool to assist in matching SmartHouse technology to the needs of the elderly and people with disabilities. The tool makes suggestions of SmartHouse devices that could assist with given impairments. SmartHouse past problem-solving textual reports have been used to obtain knowledge for the CBR system. Creating a case-based reasoning system from textual sources is challenging because it requires that the text be interpreted in a meaningful way in order to create cases that are effective in problem-solving and to be able to reasonably interpret queries. Effective case retrieval and query interpretation is only possible if a domain-specific conceptual model is available and if the different meanings that a word can take can be recognised in the text. Approaches based on methods in information retrieval require large amounts of data and typically result in knowledge-poor representations. The costs become prohibitive if an expert is engaged to manually craft cases or hand tag documents for learning. Furthermore, hierarchically structured case representations are preferred to flat-structured ones for problem-solving because they allow for comparison at different levels of specificity thus resulting in more effective retrieval than flat structured cases. This project has developed SmartCAT-T, a tool that creates knowledge-rich hierarchically structured cases from semi-structured textual reports. SmartCAT-T highlights important phrases in the textual SmartHouse problem-solving reports and uses the phrases to create a conceptual model of the domain. The model then becomes a standard structure onto which each semi-structured SmartHouse report is mapped in order to obtain the correspondingly structured case. SmartCAT-T also relies on an unsupervised methodology that recognises word synonyms in text. The methodology is used to create a uniform vocabulary for the textual reports and the resulting harmonised text is used to create the standard conceptual model of the domain. The technique is also employed in query interpretation during problem solving. SmartCAT-T does not require large sets of tagged data for learning, and the concepts in the conceptual model are interpretable, allowing for expert refinement of knowledge. Evaluation results show that the created cases contain knowledge that is useful for problem solving. An improvement in results is also observed when the text and queries are harmonised. A further evaluation highlights a high potential for the techniques developed in this research to be useful in domains other than SmartHouse. All this has been implemented in the Smarter case-based reasoning system.

This thesis focuses on the application of case based reasoning (CBR) to welding fixtures in a computer aided design (CAD) environment. Modular fixtures have become more popular in previous years due to the creation of flexible manufacturing systems. Modular fixtures, since they are composed of many standardized parts, require much iteration to produce a full fixture design. This process is made more complicated when it is applied to more complex parts such as welding assemblies. In an effort to simplify fixture design for such complicated parts, researchers have been working on integrating fixture design into CAD packages. These efforts, generally known as computer aided fixture design (CAFD), do not focus on the transition of experience from more experienced designers but only provide a structure and a virtual environment to create fixtures. The research presented in this thesis will apply to this area. Case based reasoning (CBR) is a method of using previous cases to help aid the development of solutions to new problems. Applied to CAFD, this method is reduced to the application of a database and a retrieval and adaptation system. Current research on CAFD and CBR is limited to only proposing systems for machining fixtures. This thesis presents a methodology of a CAFD and CBR system that is dedicated to welding assemblies and fixtures. The focus is on creating an indexing system that adequately represents the workpiece and fixture, a retrieval system that accurately recovers the previous cases, and a method that integrates designer feedback in each process. The results of this thesis will be shown in a case study using an automobile muffler fixture assembly to define each idea of the methodology and to provide an example.

Thesis (Doctoral)--Universität Osnabrücks, 2006. / Includes abstract and bibliographical references (p. 216-231).

The basic concept of this work is that functional modelling techniques are applicable to and of practical use in, producing a qualitative model of conceptual engineering design. A qualitative function based model of conceptual design has been developed and a computer based implementation has been built and tested. The rationale behind the modelling scheme and the computer implementation are described in detail. In addition to a review of existing models of design the research provides a significant new capability in four main areas: • An ability to generate new concepts with a controlled degree of similarity to existing designs. • A new function based model of engineering conceptual design. • The COncept Design ASsistant (CODAS) system, a computer based implementation of the function based model, has been developed and tested. • A new symbolic representation language. CODAS is a hybrid case-based and function-based modelling system, implemented in the domain of mechanical device design, which demonstrates the practical application of this new model. The CODAS system aims to provide a design support tool which can invent both routine and novel devices based on experience gained from past successful design solutions. Fast and efficient data handling is achieved by utilizing Case Based Reasoning (CBR) technology to store and retrieve past design solutions which are defined in terms of a symbolic representation language. The underlying design model is function based and employs a technique of divergent function to form mapping to produce physical embodiments of the proposed functional solutions.

No description available.

A central problem in case based reasoning (CBR) is how to store and retrieve cases. One approach to this problem is to use exemplar based models, where only the prototypical cases are stored. However, the development of an exemplar based model (EBM) requires the solution of several problems: (i) how can a EBM be represented? (ii) given a new case, how can a suitable exemplar be retrieved? (iii) what makes a good exemplar? (iv) how can an EBM be learned incrementally? This thesis develops a new model, called a probabilistic exemplar based model, that addresses these research questions. The model utilizes Bayesian networks to develop a suitable representation and uses probability theory to develop the foundations of the developed model. A probability propagation method is used to retrieve exemplars when a new case is presented and for assessing the prototypicality of an exemplar. The model learns incrementally by revising the exemplars retained and by updating the conditional probabilities required by the Bayesian network. The problem of ignorance, encountered when only a few cases have been observed, is tackled by introducing the concept of a virtual exemplar to represent all the unseen cases. The model is implemented in C and evaluated on three datasets. It is also contrasted with related work in CBR and machine learning (ML).

Thesis (Ph. D.)--University of New South Wales, 2003. / Also available online.

Case-Based Reasoning (CBR) has been successfully applied to many fields especially in the design domain. Poor assembly part design increases the cost, raises the manufacturing complexity and reduces the product quality. However, little research has been devoted to predict the potential design problems in the early design stage. The objective of this paper is to integrate the indexes of assemblability and manufacturability into adaptive phase in CBR to avoid inexperienced mistakes. Early experimental results indicate that quantitative feedback of these indexes can guide novices to depict a good assembly part design, let experienced designers confirm their experience judgments and finally impart the experience to novices through CBR methodology.

No description available.