Nuclear data evaluation: Status and perspectives

The knowledge of nuclear reaction data is an important requisite for the development and advances in nuclear technologies, but also for several fields of basic sciences such as nuclear astrophysics, nuclear medicine and material sciences. Nuclear data evaluation aims to reveal the best knowledge of nuclear reactions combining available experimental data, theoretical knowledge and mathematical constraints. A brief introduction into the concept of Bayesian evaluation techniques is given and the importance of proper inclusion of correlations between experiments is discussed. An important aspect is the generation of reliable uncertainty information which is one of the most important demands from the user community. In this context the methods for inclusion of model deficiencies are sketched. Furthermore, the specific problems of evaluations of light nuclear systems are addressed. In the outlook there will be a focus on methods and proposals taking advantage of the availability of increased storage and computational power which enable a more streamlined generation of new and/or updated nuclear data evaluations.