A list of recent papers about Graph Neural Network methods applied in NLP areas.

ConHGNN-SUM: Contextualized Heterogeneous Graph Neural Networks for extractive document summarization. Published in IEEE AISP 2024. Revolutionizes summarization by modeling documents as dynamic graphs with semantic relationships between words and sentences.

PyTorch implementation of Heterogeneous Graph Neural Networks with attention aggregation for node classification on academic networks.