Just write the list of namespaces to every split of xml dump, otherwise bliki doesn't pick up namespace (why it wouldn't look at the tag I have no idea)
Why is it important? without properly recognising wikipedia namespace, pages like this are recognised as articles and it later on skews context vectors for most of the dbpedia ids.
Closes #25.
Just write the list of namespaces to every split of xml dump, otherwise bliki doesn't pick up namespace (why it wouldn't look at the tag I have no idea)
Why is it important? without properly recognising wikipedia namespace, pages like this are recognised as articles and it later on skews context vectors for most of the dbpedia ids.