TF-IDF stands for "term frequency–inverse document frequency". It is a mathematical way of measuring how important a word is to a particular document in a collection of documents. In simple terms, it looks at how often a word appears in a document, while also taking into account how often that word appears in all the other documents. This way, words that are more unique to one document will be given more weight, while words that appear often in many documents will be given less weight. This makes it easier to tell which words a particular document relies on the most.