Learning and Linking Textual Cases

In this article, we present a new approach for the automatic acquisition and cross-linking of textual cases from heterogeneous knowledge sources like manuals, articles, and mailing lists. We develop a case generation strategy including basic transformation processes, a shallow feature extraction, and the automatic creation of cross-references. We discuss how to compute similarity and how to organise the case storage. We employ our acquisition strategy for a sample system, namely a case-based online dictionary for the RoboCup simulation league community, and show first results of our prototype.