Counting publications - Journals vs. Conferences in Computer Science

The Norwegian ranking system for publications operates with two main levels for publications in journals: level 1 for normal quality and level 2 for high-level journals. These levels are then used to award publication points to authors, and the points are increasingly used to evaluate scientists in various situations. One of the weaknesses of the system is that conference publications, which are more commonly used in some fields, are not scored on equal terms with journals. This paper highlights some of the issues with the current system and uses several metrics to compare a set of computer science conferences with journals at level 1 and 2 in the Norwegian ranking system. The comparison shows that, using these metrics, several of the conferences are at equal levels to good or high quality journals, but authors that publish at these conferences get lower scores due to weaknesses in the ranking system.