A Knowledge Recommendation Model for Power Grid Audit Based on Graph Convolutional Network
The demands for online training and auditing business of power grid enterprises are creasing,but the trainees can not effectively choose the learning objects,and this results in poor network training effect.To this end,this paper proposes a knowledge recommendation model for power grid auditing based on graph convolutional networks.An online learning knowl-edge graph is constructed based on the power grid audit business.The information in updating the graph representation is en-larged through the attention mechanism-based graph convolutional network to eliminate the influence of noise.The operation-aware network based on dense features captures implicit learners with interactive behavior,and provides them with power grid audit training knowledge recommendation.The experimental analysis shows that the model proposed has certain advantages compared with CompGCN,DeepFM and DCN methods in terms of recommendation accuracy and F1 score.
万方数据
维普
