datasets and code

[hongxiaDu]

hi, thanks for the code.
In the paper R-VGAE, Sime-supervised uses concept-concept edges, but in the code there is only document-concept, document-document adjacency matrix.
if args.ds.startswith('tf'): if args.labels == 'y': adj_cd, adj_dd, features, tags_nodes = my_load_data_tfidf_semi(args.wmd) else: adj_cd, adj_dd, features = my_load_data_tfidf(args.wmd)

1. I want to know where concept-concept edge is used?
2. When using TF-IDF as the embedding feature, how is the feature of the concept obtained?
3. I know the tags of concepts are 0-321, what are the tags of documents?