Natural language processing
NLP is a subfield of machine learning that allows computers to interpret, manipulate, and comprehend human language. This definition sounds a little dry, so, to provide a little clarity, let’s start with a non-exhaustive list of the types of tasks that fall under the NLP umbrella:
- Text classification: This assigns a single label to the entire input text. For example, sentiment analysis can determine whether a product review is positive or negative.
- Token classification: This assigns a label for each token of the input text. A token is a building block (or a unit) of text. Words can be tokens. A popular token classification task is named entity recognition, which assigns each token to a list of predefined classes such as place, company, or person. Part-of-speech (POS) tagging assigns each word to a particular part of speech, such as a noun, verb, or adjective.
- Text generation: This uses the input text to generate new text with arbitrary...
