You're reading from Mastering spaCy Build structured NLP solutions with custom components and models powered by spacy-llm

Product type Paperback

Published in Feb 2025

Publisher Packt

ISBN-13 9781835880463

Length 238 pages

Edition 2nd Edition

Languages

Python

Tools

Spacy

Concepts

GPT/LLMs

Authors (2):

Déborah Mesquita

Duygu Altınok

View More author details

Table of Contents (17) Chapters

Preface

1. Part 1: Getting Started with spaCy

2. Chapter 1: Getting Started with spaCy FREE CHAPTER

3. Chapter 2: Core Operations with spaCy

4. Part 2: Advanced Linguistic and Semantic Analysis

5. Chapter 3: Extracting Linguistic Features

6. Chapter 4: Mastering Rule-Based Matching

7. Chapter 5: Extracting Semantic Representations with spaCy Pipelines

8. Chapter 6: Utilizing spaCy with Transformers

9. Part 3: Customizing and Integrating NLP Workflows

10. Chapter 7: Enhancing NLP Tasks Using LLMs with spacy-llm

11. Chapter 8: Training an NER Component with Your Own Data

12. Chapter 9: Creating End-to-End spaCy Workflows with Weasel

13. Chapter 10: Training an Entity Linker Model with spaCy

14. Chapter 11: Integrating spaCy with Third-Party Libraries

15. Index

Why subscribe?

16. Other Books You May Enjoy

Creating patterns with PhraseMatcher

While processing financial, medical, or legal text, we often have long lists and dictionaries and we want to scan the text against our lists. As we saw in the previous section, Matcher patterns are quite handcrafted; we coded each token individually. If you have a long list of phrases, Matcher is not very handy. It’s not possible to code all the terms one by one.

spaCy offers a solution for comparing text against long dictionaries – the PhraseMatcher class. The PhraseMatcher class helps us match long dictionaries. Let’s get started with a basic example of using PhraseMatcher to match terms defined in a list:

Import the library and the class and instantiate the nlp pipeline as usual:

import spacy
from spacy.matcher import PhraseMatcher
nlp = spacy.load("en_core_web_sm")

Now we can instantiate the PhraseMatcher object and call nlp.make_doc() on the terms one by one to create patterns:
```
matcher = PhraseMatcher...
```