We present a formalism and a method that allow to learn structured representations in a noisy knowledge base. The formalism fills the gap between Artificial Intelligence and Data Analysis research's domains. It describes a kind of structured modal object called “hoard”, that can express various semantics as probability, possibility and belief. The method aims at growing and refining a knowledge base, composed of hoards, by the incremental use of a data base, composed of hoards subparts. Two levels of knowledge are involved: structured hoards represent classes, and non structured individuals represent data. The main goals of the method are to match quickly both levels by automatic rules generation, and to acquire new knowledge in the structured level, by using the flat data base. Some applications are, for example, situation understanding, image analysis, adaptation to temporal variable process, negotiation.

We propose in this paper a general framework for integrating inductive and case-based reasoning (CBR) techniques for diagnosis tasks. We present a set of practical integrated approaches realised between the Kate-Induction decision tree builder and the Patdex case-based reasoning system. The integration is based on the deep understanding about the weak and strong points of each technology. This theoretical knowledge permits to specify the structural possibilities of a sound integration between the relevant components of each approach. We define different levels of integration called “cooperative”, “workbench” and “seamless”. They realise respectively a tight, medium and strong link between both techniques. Experimental results show the appropriateness of these integrated approaches for the treatment of noisy or unknown data.

This paper focuses on integrating inductive inference and case-based reasoning. We study integration along two dimensions: Integration of case-based methods with methods based on general domain knowledge, and integration of problem solving and incremental learning from experience. In the Inreca system, we perform case-based reasoning as well as tdidt (Top-Down Induction of Decision Trees) classification by using the same data structure called the Inreca tree. We extract decision knowledge using a tdidt algorithm to improve both the similarity assessment by determining optimal weights, and the speed of the overall system by inductive learning. The integrated system we implemented evolves smoothly along application development time from a pure case-based reasoning approach, where each particular case is a piece of knowledge, to a more inductive approach where some subsets of the cases are generalised into abstract knowledge. Our proposed approach is driven by the needs of a concrete pre-commercial system and real diagnostic applications. We evaluate the system on a database of insurance risk for cars and an application involving forestry management in Ireland.

An industrial strength hypermedia system permits users to have access to a wide range of stored information, in a fully structured manner. The paper demonstrates that, by using an open hypermedia system together with case-based reasoning, the details of which are discussed in the paper, significant benefits can be achieved in maintenance provision. This approach to information management is illustrated by the development of an information system for maintenance provision for a large welding cell at the Odense Steel Shipyard Ltd. The paper concludes by identifying future developments, including advanced user interfaces, access to remote information resources, and real time diagnostics, that will further enhance this approach to information provision within the industrial environment

Reliable and effective maintenance support is a vital consideration for the management within today’s manufacturing environment. This paper discusses the development a maintenance system for the world largest robot welding facility. The developed system combines a case-based reasoning approach for diagnosis with context information, as electronic on-line manuals, linked using open hypermedia technology. The work discussed in this paper delivers not only a maintenance system for the robot stations under consideration, but also a design framework for developing maintenance systems for other similar applications.

In this chapter, we will focus on the utilization of Case-Based Reasoning for solving problems in the area of diagnosis and decision support. For this, we will first discuss different types of analytic problem solving, explain alternative approaches of coping with specific problems, and finally sketch a number of successful applications.

this paper we focus on a deep integration ofthese technologies.First, we motivate our approach on a more intuitive basis. Chapter 2 results in amore or less concrete guideline for integrating inductive and case-based reasoningbased on mechanisms known from the field of information retrieval. We introducemultidimensional retrieval structures for associative search, especially k-d trees anddescribe the basic algorithms for tree construction and search. These basic datastructures and...

12 generated. Such a hypothesis consists of a pair of a set of cases and an associated similarity measure. Questions to be answered are: Which cases will be entered into the case base, which will be removed from it, and how to realise the similaritymeasure. Since the criteria of Jantke and Lange can be analogously applied to case-based learning, it can be viewed as a special instance of inductive learning. A common theoretical framework is necessary to achieve results on this topic. First steps in this direction have been described by Jantke, Richter et al. Both inductive learning and case-based learning have in common that they derive "global" knowledge from "local" observations (which, of course, are uncertain, respectively). However, they use diierent techniques to achieve this: Inductive learning bases mainly on logical concept descriptions ("logical reasoning"), whereas case-based reasoners often use analytic descriptions ("geometric reasoning") (cf., e.g., Richter, 1992). One consequence from this is that inductive learners mostly start with the "dropping of complete dimensions" in contrast to case-based reasoners which "de-compose complete dimensions into intervals". It depends on the use of a learning result which particular technique is then the more successful one. Therefore, the Inreca approach integrates both learning strategies within a broader architecture for identiication and diagnostic reasoning. Up to now, much work has been done on the integration of diierent knowledge representation and processing schemes to improve knowledge acquisition. E.g., a comparative analysis as well as a proposed integration of models, cases and compiled knowledge have been given by van Someren, Zheng and Post (1990). The Moltke architecture also bases on these three schemes (cf. The Granul system integrates several existing knowledge acquisition tools into one coherent system that supports several styles of knowledge acquisition (Aben, van Someren & Terpstra, 1992). The Mobal system is an interesting example for the integration of manual and automatic knowledge acquisition methods (the balanced cooperative modelling issue, cf. Morik, 1991). Van de Velde and Aamodt (1992) have analysed the possible use of machine learning techniques within the Kads approach to expert system development. Rissland and Skalag (1989) introduced the notion of mixed paradigm reasoning for the integration of diierent reasoning schemes (reasoning from cases, rules, constraints, deep models etc.). others. A rst suggestion for the integration of case-based reasoning and model-based knowledge acquisition is given in Janetzko and Strube (1992). 7 Conclusion We have introduced basic parts of the architecture …