Pascal Reuss

Relevance Matrix Generation Using Sensitivity Analysis in a Case-Based Reasoning Environment

Relevance matrices are a way to formalize the contribution of each attribute in a classification task. Within the CBR paradigm these matrices can be used to improve the global similarity function that outputs the similarity degree of two cases, which helps facilitate retrieval. In this work a sensitivity analysis method was developed to optimize the relevance values of each attribute of a case in a CBR environment, thus allowing an improved comparison of cases. The process begins with a statistical analysis of the values in a given dataset, and continues with an incremental update of the relevance of each attribute.

Semi-automatic Knowledge Extraction from Semi-structured and Unstructured Data Within the OMAHA Project

This paper describes a workflow for semi-automatic knowledge extraction for case-based diagnosis in the aircraft domain. There are different types of data sources: structured, semi-structured and unstructured source. Because of the high number of data sources available and necessary, a semi-automatic extraction and transformation of the knowledge is required to support the knowledge engineers. This support shall be performed by a part of our multi-agent system for aircraft diagnosis. First we describe our multi-agent system to show the context of the knowledge extraction. Then we describe our idea of the workflow with its single tasks and substeps. At last the current implementation, and evaluation of our system is described.

FEATURE-TAK - Framework for Extraction, Analysis, and Transformation of Unstructured Textual Aircraft Knowledge

This paper describes a framework for semi-automatic knowledge extraction for case-based diagnosis in the aircraft domain. The available data on historical problems and their solutions contain structured and unstructured data. To transform these data into knowledge for CBR systems, methods and algorithms from natural language processing and case-based reasoning are required. Our framework integrates different algorithms and methods to transform the available data into knowledge for vocabulary, similarity measures, and cases. We describe the idea of the framework as well as the different tasks for knowledge analysis, extraction, and transformation. In addition, we give an overview of the current implementation, our evaluation in the application context, and future work.

Weighted One Mode Projection of a Bipartite Graph as a Local Similarity Measure.

Bipartite graphs are a common structure to model relationships between two populations. Many times a compression of the graph to one population, namely a one mode projection (OMP), is needed in order to gain insight into one of the populations. Since this compression leads to loss of information, several works in the past attempted to quantify the connection quality between the items from the population that is being projected, but have ignored the edge weights in the bipartite graph. This paper presents a novel method to create a weighted OMP (WOMP) by taking edge weights of the bipartite graph into account. The usefulness of the method is then displayed in a case-based reasoning (CBR) environment as a local similarity measure between unordered symbols, in an attempt to solve the long-tail problem of infrequently used but significant symbols of textual CBR. It is shown that our method is superior to other similarity options.

Dependency Modeling for Knowledge Maintenance in Distributed CBR Systems

Knowledge-intensive software systems have to be continuously maintained to avoid inconsistent or false knowledge and preserve the problem solving competence, efficiency, and effectiveness. The more knowledge a system contains, the more dependencies between the different knowledge items may exist. Especially for an overall system, where several CBR systems are used as knowledge sources, several dependencies exist between the knowledge containers of the CBR systems. The dependencies have to be considered when maintaining the CBR systems to avoid inconsistencies between the knowledge containers. This paper gives an overview and formal definition of these maintenance dependencies. In addition, a first version of an algorithm to identify these dependencies automatically is presented. Furthermore, we describe the current implementation of dependency modeling in the open source tool myCBR.

Knowledge Engineering for Decision Support on Diagnosis and Maintenance in the Aircraft Domain

The diagnosis of machines belonging to technical domains demands careful attention: dozens of relations between individual parts have to be considered and operational or environmental conditions can effect measurable symptoms and diagnoses. An aircraft is one of the most complex machines built by humans and therefore the diagnosis and maintenance of aircrafts requires intelligent and efficient solutions.

Dynamic Case Bases and the Asymmetrical Weighted One-Mode Projection

Building a case base for a case-based reasoning (CBR) system is incomplete without similarity measures. For the attribute-value case structure similarity between values of an attribute should logically fit their relationship. Bipartite graphs have been shown to be a good representation of relationships between values of symbolic attributes and the diagnosis of the cases in a technical diagnosis CBR system, while using an asymmetrical weighted one-mode projection on the values to model their similarity.