Wrap-up

Natural Language Processing with spaCy

Azadeh Mobasher

Principal data scientist

Chapter 1 - Introduction to NLP and spaCy

spaCy Language pipeline

Work with spaCy's classes such as Doc, Token and Span and predict semantic similarities using word vectors:

Analogies and vector operations

Write matching patterns to extract terms and phrases using spaCy's Matcher and PhraseMatcher:

matcher = Matcher(nlp.vocab)
pattern = [{"LOWER": "good"}, {"LOWER": {"IN": ["morning", "evening"]}}]
matcher.add("morning_greeting", [pattern])

matcher = PhraseMatcher(nlp.vocab, attr = "LOWER")
patterns = [nlp.make_doc(term) for term in terms]
matcher.add("InvestmentTerms", patterns)

Example of a medical domain NER

Natural Language Processing with spaCy