Abstract: In order to realize efficient continuous graph query, a dynamic graph query method based on Top-k continuous matching mechanism is proposed. Continuous matching was identified by traversing the marked version graph, in which each node, edge and label existed in the interval set in the graph. The efficient memory layout of the labeled version graph allowed fast retrieval of adjacent nodes at each snapshot. In order to reduce the number of candidate matches, the neighborhood and path time index based on Bloom filter were further introduced, and the duration threshold was determined by using the time index. The experimental results show that the proposed method can effectively realize the most sustainable dynamic graph query, and has high computational efficiency.