Reputation: 21
Given a text, I am able to recognise all entities by using a discriminative NER model. Let us imagine that we have a page of text in which the different entities that make up a group of people are described and for each of them we have data such as 'first name', 'surname' and 'date of birth'. The NER model is able to find all these entities but is unable to group them into the 'person' object. How could I solve this NLP task?
I have tried to group these entities based on the span in the original text using heuristics. but this requires a lot of effort for more complex cases. Also, if these data are extracted from different pages?
Upvotes: 2
Views: 77