Wrapping up

Advanced NLP with spaCy

Ines Montani

spaCy core developer

Your new spaCy skills

• Extract linguistic features: part-of-speech tags, dependencies, named entities
• Work with pre-trained statistical models
• Find words and phrases using Matcher and PhraseMatcher match rules
• Best practices for working with data structures Doc, Token Span, Vocab, Lexeme
• Find semantic similarities using word vectors
• Write custom pipeline components with extension attributes
• Scale up your spaCy pipelines and make them fast
• Create training data for spaCy' statistical models
• Train and update spaCy's neural network models with new data

More things to do with spaCy (1)

• Training and updating other pipeline components
  • Part-of-speech tagger
  • Dependency parser
  • Text classifier

More things to do with spaCy (2)

• Customizing the tokenizer
  • Adding rules and exceptions to split text differently
• Adding or improving support for other languages
  • 45+ languages currently
  • Lots of room for improvement and more languages
  • Allows training models for other languages

See the website for more info and documentation!

spacy.io

Thanks and see you soon!

Advanced NLP with spaCy