In a lot of case we will have a lot of tsv files with the same columns, we want to be able to get all the values for a predicate without needing to filter on every file so it would be better to have the following predicates:
http://data2services/my_column1
If we want to filter on file (or table) we can use the rdf:type
rr:predicate http://data2services/data/my_file.tsv/my_column1
In a lot of case we will have a lot of tsv files with the same columns, we want to be able to get all the values for a predicate without needing to filter on every file so it would be better to have the following predicates: http://data2services/my_column1
If we want to filter on file (or table) we can use the rdf:type