Chapter 5: Getting Started with Information Extraction
In this chapter, we will cover the basics of information extraction. We will start with extracting emails and URLs from job announcements. Then we will use an algorithm called the Levenshtein distance to find similar strings. Next, we will use spaCy to find named entities in text, and later we will train our own named entity recognition (NER) model in spaCy. We will then do basic sentiment analysis, and finally, we will train two custom sentiment analysis models.
You will learn how to use existing tools and train your own models for information extraction tasks.
We will cover the following recipes in this chapter:
- Using regular expressions
- Finding similar strings: the Levenshtein distance
- Performing NER using spaCy
- Training your own NER model with spaCy
- Discovering sentiment analysis
- Sentiment for short texts using LSTM: Twitter
- Using BERT for sentiment analysis
