Can Persistent Homology provide an efficient alternative for Evaluation of Knowledge Graph Completion Methods?