GOLEM-lab / fandom-wiki

Extraction of structured and unstructured information from fandom.com pages
10 stars 0 forks source link

Implement NER and POS #3

Open txetxedeletxe opened 1 year ago

txetxedeletxe commented 1 year ago

The current system uses entities from article names / infobox as subjects of the relations, and looks for objects for a list of target relations. This potentially limits the relations that can be extracted from the text, implementing NER (Name Entity Recognition) and POS classification will allow to capture these relations.

txetxedeletxe commented 1 year ago

This is not super important as the primary use-case consists in getting relations for a character given its fandom article, thus the subject of the relation is already given in those cases.