Evaluation of query languages with software science metrics

Abstract

Current methods of evaluating user's difficulties with query languages through empirical studies are very tedious. The empirical results have very little validity beyond the few queries tested in the studies. To allow a quicker and more wide-ranging evaluation, we apply Software Science Metrics (SSM) to query languages. The results from the metrics analyses are compared with results from experimental studies. Statistical analyses produce very good models to predict query accuracy. The results indicate that SSM is a good alternative for the evaluation of query languages. The results can be used in the evaluation and design of query languages.