notebook.community

Edit and run



In [45]:

    
import spacy
from spacy import displacy



In [46]:

    
nlp = spacy.load('en')



In [47]:

    
def display_entities(doc):
    if doc.ents:
        for ents in doc.ents:
            print(ents.text," - ",ents.label_," - ",str(spacy.explain(ents.label_)))
    else:
        print("No entities found in the document")



In [48]:

    
doc1 = nlp(u"Hi How are you")

display_entities(doc1)









    



No entities found in the document



In [61]:

    
doc2 = nlp(u"Welcome to London. Hope you have a good time Mr. Anderson. When is your next flight to Bombay? Could you come over sometime to Zurich / Swiss")

display_entities(doc2)









    



London  -  GPE  -  Countries, cities, states
Anderson  -  PERSON  -  People, including fictional
Bombay  -  GPE  -  Countries, cities, states
Zurich  -  GPE  -  Countries, cities, states



In [62]:

    
displacy.render(doc2,style='ent',jupyter=True)









    




Welcome to 

    London
    GPE

. Hope you have a good time Mr. 

    Anderson
    PERSON

. When is your next flight to 

    Bombay
    GPE

? Could you come over sometime to 

    Zurich
    GPE

 / Swiss



In [51]:

    
doc3 = nlp(u"Can i borrow 1000 rupees from you? or you can give me 100 dollars of Microsoft stocks")

display_entities(doc3)









    



1000  -  CARDINAL  -  Numerals that do not fall under another type
100 dollars  -  MONEY  -  Monetary values, including unit
Microsoft  -  ORG  -  Companies, agencies, institutions, etc.



In [52]:

    
displacy.render(doc3,style='ent',jupyter=True)









    




Can i borrow 

    1000
    CARDINAL

 rupees from you? or you can give me 

    100 dollars
    MONEY

 of 

    Microsoft
    ORG

 stocks



In [53]:

    
doc4 = nlp(u'Tesla to build a U.K. factory for $6 million')

display_entities(doc4)  ## note : Tesla is not recorgnised









    



U.K.  -  GPE  -  Countries, cities, states
$6 million  -  MONEY  -  Monetary values, including unit

Adding missing tokens as named-entities



In [54]:

    
from spacy.tokens import Span

ORG = doc4.vocab.strings[u'ORG']  ## returns id of ORG label
print(ORG)

new_ent = Span(doc4,0,1,label=ORG)

## add entity to an existing doc
doc4.ents = list(doc4.ents) + [new_ent]  ## note you can run only once



In [55]:

    
display_entities(doc4)









    



Tesla  -  ORG  -  Companies, agencies, institutions, etc.
U.K.  -  GPE  -  Countries, cities, states
$6 million  -  MONEY  -  Monetary values, including unit



In [56]:

    
displacy.render(doc4,'ent', jupyter=True)









    






    Tesla
    ORG

 to build a 

    U.K.
    GPE

 factory for 

    $6 million
    MONEY



In [ ]: