Finding Explanations in Textual Reports

The paper presents our work on extracting explanations from textual reports which then are used in a CaseBased Reasoning (CBR) system to support the suggested solutions and creating trust in the system. We describe how explanations are extracted from text and represented in a structured representation, called Text Reasoning Graph. Our hypothesis is that the extracted explanations provide an explicit connection between a description of a problem and its solution. In this work we introduce an automatic evaluation measure for explanations to verify this hypothesis.