A Quantum Spatial Graph Convolutional Network for Text Classification

• DOI: 10.32604/csse.2021.014234
• Publication Date: 2021-01-05T10:35:55Z
• Authors: Syed Mustajar Ahm...; Hongwei Ge; Sami Ahmed Haider; Muhammad Irshad; Sohail M. Noman; Jehangir Arshad; Asfandeyar Ahmad; Talha Younas
• Keywords: Adjacency matrix

The data generated from non-Euclidean domains and its graphical representation (with complex-relationship object interdependence) applications has observed an exponential growth. sophistication of graph posed consequential obstacles to the existing machine learning algorithms. In this study, we have considered a revamped version semi-supervised algorithm for graph-structured address issue expanding deep approaches represent data. Additionally, quantum information theory been applied through Graph Neural Networks (GNNs) generate Riemannian metrics in closed-form several layers. further, pre-process adjacency matrix graphs, new formulation is established incorporate high order proximities. proposed scheme shown outstanding improvements overcome deficiencies Convolutional Network (GCN), particularly, loss imprecise with acceptable computational overhead. Moreover, Quantum (QGCN) significantly strengthened GCN on node classification tasks. parallel, it expands generalization process significant difference by making small random perturbations during training process. evaluation results are provided three benchmark datasets, including Citeseer, Cora, PubMed, that distinctly delineate superiority model terms accuracy against state-of-the-art other methods based same algorithms literature.