Attributed relational graph matching neural network and its application

A new method of error-calibrated and attributed relational matching neural network (EARGMNN) was developed in this paper. In attributed relational graphs (ARG), there are direction arcs and multi-arcs, so ARG is asymmetric, but the Hopfield net is symmetry. After redefining the distances of node feature and node-relational arc feature, these asymmetry problem were solved. At the same time, the idea of error-calibration was introduced into neural network. Then the net can be used as random semantic net matching. The analogue annealing method was introduced in EARGMNN model also, the test results are quite satisfactory.