POS Tagging with spaCy in NLP: Simple Guide and Example

To do POS tagging with spaCy in NLP, load a language model like en_core_web_sm, process your text with nlp(), and access the pos_ attribute of each token for its part-of-speech tag. This gives you easy access to grammatical tags for each word in your text.

Syntax

Use spacy.load() to load a language model. Then call nlp(text) to process your text. Each word becomes a Token object. Access token.pos_ for the POS tag as a readable string.

• spacy.load('en_core_web_sm'): loads English model
• nlp(text): processes text into tokens
• token.pos_: gets POS tag of a token

python

import spacy

nlp = spacy.load('en_core_web_sm')
doc = nlp('This is a sentence.')
for token in doc:
    print(token.text, token.pos_)

Output

This DET
is AUX
a DET
sentence NOUN
. PUNCT

Example

This example shows how to load spaCy's English model, process a sentence, and print each word with its POS tag.

python

import spacy

# Load the small English model
nlp = spacy.load('en_core_web_sm')

# Text to analyze
text = 'SpaCy is a great library for NLP tasks.'

# Process the text
doc = nlp(text)

# Print each token with its POS tag
for token in doc:
    print(f'{token.text}: {token.pos_}')

Output

SpaCy: PROPN
is: AUX
a: DET
great: ADJ
library: NOUN
for: ADP
NLP: PROPN
tasks: NOUN
.: PUNCT

Common Pitfalls

Common mistakes include not loading a model before processing text, which causes errors, or confusing pos_ with tag_. The pos_ attribute gives coarse-grained POS tags (like NOUN, VERB), while tag_ gives fine-grained tags (like NN, VBD).

Also, forgetting to install the model with python -m spacy download en_core_web_sm leads to loading errors.

python

import spacy

# Wrong: not loading model
# nlp = None
# doc = nlp('Test sentence')  # This will raise an error

# Right way:
nlp = spacy.load('en_core_web_sm')
doc = nlp('Test sentence')
for token in doc:
    print(token.text, token.pos_)

Output

Test NOUN
sentence NOUN

Quick Reference

Attribute	Description	Example Value
nlp = spacy.load('en_core_web_sm')	Load English model	nlp object
doc = nlp(text)	Process text into tokens	doc object
token.text	Original word text	'SpaCy'
token.pos_	Coarse POS tag	'NOUN', 'VERB', 'ADJ'
token.tag_	Fine-grained POS tag	'NN', 'VBD', 'JJ'

Key Takeaways

Always load a spaCy language model before processing text for POS tagging.

Use token.pos_ to get easy-to-understand part-of-speech tags for each word.

Install models with 'python -m spacy download en_core_web_sm' if not already installed.

Remember pos_ gives coarse tags; tag_ gives more detailed POS tags.

Process text with nlp(text) to get tokens for POS tagging.