I am trying to improve the result in an experiment that evaluate the performance of the concatenation between contextualized word embedding and Knowledge Graph Embedding, co
