It is difficult to analyze and understand 3D data for machine interpretation. In this work, we suggest utilizing Graph Matching (GM) for sign identification based on 3D motion collected ISL data. We formulate the sign identification and categorization issue using 3D motion signs as an AGM. However, there are significant limitations with the present models, the most notable of which are (1) spatial matching on a defined set of frames with fixed nodes and (2) temporal match ing breaking the full 3D dataset into fixed pyramids. These issues are addressed by our method, which employs spatial matching between frames and temporal matching for numerous intra-frame matches. A 3D sign language dataset consisting of 200 continuous phrase signs was captured using an 8-camera motion capture setup to evaluate the proposed model. We demonstrate that our method improves the reliability of sign recognition in ongoing sign language discourse.