Things to change in generator for different dataset

DeNederlandscheBank / nqm

A Transformer-based Machine for answering questions on insurance companies

MIT License

0 stars 0 forks source link

Things to change in generator for different dataset #4

Closed jm-glowienke closed 3 years ago

jm-glowienke commented 3 years ago

At the moment the generator is specific for DBpedia. Accordingly, some things have to be adapted to make the generator working with a different dataset:

[x] Likely: String replacements for ontolgy classes and LABEL_REPLACEMENT (not 100% sure to which extent)
[x] query_dbpedia function (for sure)
[x] normalize(ontology_class)
[ ] prioritize_usage(match) might need to be adapted (depends on size of dataset, accordingly usage preference has to be adapted)
[ ] REPLACEMENTS
[ ] STANDARDS

jm-glowienke commented 3 years ago

The generator module is working for the EIOPA register data. Some adjustments from above were not necessary (yet).

For now, this issue can be closed. Some remaining open tasks are noted down on the project board.