AAE: An active auto-estimator for improving graph storage

Graph has currently become an increasingly popular model in many real applications. The efficiency of graph storage is crucial for these Generally speaking, the tuning task relies on database administrators to find best storage. However, DBAs make decisions by mainly relying their experiences and intuition. Due limitations DBAs, may have uncertain performance worse efficiency. In this paper, we observed that estimator workload potential ability guarantee operations. Unfortunately, because complex characteristics evaluation task, there exists no mature workload. We formulate as a classification carefully design feature engineering process, including data features, features features. For propose active auto-estimator (AAE) combining learning deep learning. test time accuracy AAE with two open source data, LDBC Freebase. experimental results show our could efficiently complete milliseconds.