We need to add an additional phase to our pipeline were we will extract the named entities from a given article. These entities will be later on mapped to Wikidata, and a list of topics will be inferred from their class and neighbourhood.
For now, I think we could try out the default Spacy Entity Recognizer and see if the results we obtain are satisfactory enough to be used in the later phases.
We need to add an additional phase to our pipeline were we will extract the named entities from a given article. These entities will be later on mapped to Wikidata, and a list of topics will be inferred from their class and neighbourhood.
For now, I think we could try out the default Spacy Entity Recognizer and see if the results we obtain are satisfactory enough to be used in the later phases.