Linguistic features

Natural Language Processing with spaCy

Azadeh Mobasher

Principal Data Scientist

POS tagging

POS tags depend on the context, surrounding words and their tags

import spacy
nlp = spacy.load("en_core_web_sm")
text = "My cat will fish for a fish tomorrrow in a fishy way."
print([(token.text, token.pos_, spacy.explain(token.pos_)) 
        for token in nlp(text)])

POS tagging example

What is the importance of POS?

Better accuracy for many NLP tasks

I will fish tomorrow.

I ate fish.

Translation system use case

verb -> pescaré

noun -> pescado

What is the importance of POS?

Word-sense disambiguation (WSD) is the problem of deciding in which sense a word is used in a sentence.
Determining the sense of the word can be crucial in machine translation, etc.

WSD Example

Word-sense disambiguation

import spacy
nlp = spacy.load("en_core_web_sm")

verb_text = "I will fish tomorrow."
noun_text = "I ate fish."


print([(token.text, token.pos_) for token in nlp(verb_text) if "fish" in token.text], "\n")
print([(token.text, token.pos_) for token in nlp(noun_text) if "fish" in token.text])

[('fish', 'VERB', 'verb')] 
[('fish', 'NOUN', 'noun')]

Dependency parsing

Explores a sentence syntax
Links between two tokens
Results in a tree

Example of dependency parsing

Dependency parsing and spaCy

Dependency label describes the type of syntactic relation between two tokens

Dependency label	Description
nsubj	Nominal subject
root	Root
det	Determiner
dobj	Direct object
aux	Auxiliary

Dependency parsing and displaCy

displaCy can draw dependency trees

doc = nlp("We understand the differences.")

spacy.displacy.serve(doc, style="dep")

Example of dependency parser with displaCy

Dependency parsing and spaCy

.dep_ attribute to access the dependency label of a token

doc = nlp("We understand the differences.")
print([(token.text, token.dep_, spacy.explain(token.dep_)) for token in doc])

[('We', 'nsubj', 'nominal subject'), ('understand', 'ROOT', 'root'),
('the', 'det', 'determiner'), ('differences', 'dobj', 'direct object'),
('.', 'punct', 'punctuation')]

Let's practice!

Natural Language Processing with spaCy