Mining Automotive Warranty Claims Data for Effective Root Cause Analysis

We present an application of text analytics in automotive industry and describe a research prototype for extracting named-entities in textual data recorded in automotive warranty claim forms. We describe an application for gaining useful insights about products defect reported to the dealer during the warranty period of vehicles. The prototype is developed for air-conditioning subsystem and consists of two main components: a text tagging and annotation engine a query engine. We present some real world examples with sample output and share our design and implementation experiences.