In recent years manufacturing enterprises are increasingly automated and collect and store large quantities of data relating to their products and production systems. This electronically stored data can hold both process measures and hidden information, which can be very important when discovered. Knowledge discovery in databases provides the tools to explore historic or current data to reveal many kinds of previously unknown knowledge from these databases. Manufacturing enterprises data is complex and may include information relating to design, process improvement and limitations, manufacturing machines and tools, and product quality. This thesis focuses on issues relating to information extraction from engineering databases in general and from manufacturing processes in particular using their historical databases. It also addresses the important issue of how the process or the design of the product can be improved based on such information.