Part-of-Speech Tagging

Introducing POS Tagging

• After the tokenization process is completed and transforms the raw text into a Doc object, spacy can perform POS tagging on that Doc object
• This is where the statistical model comes into play
• The statistical model is used to make predictions about which tags most likely apply to a token in the given context

The Algorithm for Tagging

1. The Tagger receives a Doc object.
2. The Tagger iterates over the Token objects.
3. The Tagger assigns each Token object a tag attribute based on a CNN.

References

• Part-of-Speech Tagging in Spacy
• Tagger Class