EntityRuler
While covering Matcher, we saw that we can extract named entities with Matcher by using the ENT_TYPE attribute. We recall from the previous chapter that ENT_TYPE is a linguistic attribute that refers to the entity type of the token, such as person, place, or organization. Let's see an example:
pattern = [{"ENT_TYPE": "PERSON"}]
matcher.add("personEnt", [pattern])
doc = nlp("Bill Gates visited Berlin.")
matches = matcher(doc)
for mid, start, end in matches:
print(start, end, doc[start:end])
...
0 1 Bill
1 2 Gates
Again, we created a Matcher object called matcher and called it on the Doc object, doc. The result is two tokens, Bill and Gates; Matcher always matches at the token level. We got Bill and Gates, instead of the full entity, Bill Gates. If you want to get the full entity rather than the individual tokens, you can do this:
pattern = [{"ENT_TYPE": "PERSON"...
