Uncertainty Annotated Databases - A Lightweight Approach for Approximating Certain Answers

摘要: Certain answers are a principled method for coping with uncertainty that arises in many practical data management tasks. Unfortunately, this is expensive and may ex- clude useful (if uncertain) answers. Thus, users frequently resort to less approaches resolve uncertainty. In paper, we propose Uncertainty Annotated Databases (UA-DBs), which combine an under- over-approximation of certain achieve the reliability answers, performance classical database system. Furthermore, contrast prior work on UA-DBs higher utility by including some (explicitly marked) not certain. based incomplete K-relations, introduce generalize set-based notion databases much larger class models. Using implementation our approach, demonstrate experimentally it efficiently produces tight approximations high utility.