Comparison of Structured vs. Unstructured Data for Industrial Quality Analysis

Industrial methods for quality analysis massively rely on structured data describing product features and product usage. The analysis of such data is normally done using complex reporting or sophisticated data mining methods. Besides this structured data, companies very often also posses large amounts of unstructured text like call center reports, internet fora or repair order documents. Despite the rising interest in text mining applications for industrial usage, the uncertainty about the real benefits is still high. In this work, we will present a comparison of the usage of structured versus unstructured data on two quality analysis use cases: Early warning and the detection of repeat repairs.