The invention discloses a data processing method and a related device. The method comprises: acquiring a to-be-processed adjacency graph, n nodes in the to-be-processed adjacency graph corresponding to to-be-processed data respectively, n being a positive integer, the n nodes comprising to-be-labeled nodes, and the to-be-processed data corresponding to the to-be-labeled nodes being to-be-labeled data; based on the annotation data of the annotated nodes in the n nodes, using a graph convolution network to carry out multiple times of iterative processing on the to-be-processed adjacent graph toobtain a classification result of the to-be-annotated nodes in the to-be-processed adjacent graph; and determining annotation data of the to-be-annotated node based on a classification result of the to-be-annotated node obtained by at least one time of iteration processing in the multiple times of iteration processing. The invention further discloses a corresponding device. Therefore, the to-be-labeled data in the to-be-processed data is labeled.