This project is built on top of the paper "In-Memory Subgraph Matching: an In-depth Study" by Dr. Shixuan Sun and Prof. Qiong Luo. https://github.com/RapidsAtHKUST/SubgraphMatching. We have improved the pruning power of the local filters by an additional spectral filter utilizing structure information of vertices. The result shows the spectral filter can reduce the query evaluation time in the scenario where the label information of the dataset is limited. For more detail please read my master's thesis here.