Challenges for semi-automatic trace recovery in the automotive domain

Abstract

This position paper describes the challenges of semi-automatically recovering traceability information in specifications of the automotive domain using information retrieval techniques. Recovering traceability information itself is not an easy task and a combination of factors make this task even harder in the automotive domain. For example: There are huge specifications compared to specifications evaluated in research. The specifications are written in German instead of the English generally reported, which makes it even more complicated. Improving the applied algorithms is one way of tackling these obstacles, better guidance on how and when to employ the semi-automatic recovery methods also play an important role. An approach to tackle these obstacles as well as preliminary findings are described.