chocowu / esgcn-aop Goto Github PK

Hi,dear author.
I have recently read your paper "High-Order Pair-Wise Aspect and Opinion Terms Extraction With Edge-Enhanced Syntactic Graph Convolution". I'm very interested in this article and have studied it carefully. In order to further understand the idea of the paper, I'm wondering if you could kindly send me the config file about training. Because I try to reproduce your result, I only get the 65.89 f1 in 14lap dataset. I think it might be because the configuration is different from yours. I promise it will be used only for research purposes.
Thank you very much for your kind consideration and I am looking forward to your early reply.