I want to ask that where do the TD-IDF of a word comes from? I guess maybe it is calculated in another dataset.