Natural Language Processing and Computational Linguistics

Language models

One of spaCy's most interesting features is its language models [15]. A language model is a statistical model that lets us perform the NLP tasks we want to, such as POS-tagging and NER-tagging. These language models do not come packaged with spaCy, but need to be downloaded – we'll get into details of exactly how we can download these models later on in the chapter.

Different languages have different models to perform these tasks, and there are also different models for the same language – the difference between these models is mostly statistical, and you can use different models based on your use case. A different model would just be trained on a different dataset. It is still the same underlying algorithm. The spaCy documentation ...