Indexing 2: inverted index 

Thanks!

I think that it's called inverted because they were originally created as number of occurrence of each word in a document, but at last after constructing a matrix of multiple documents, instead of using the rows in the matrix (number of occurrence of each word in the document) we use the column (number of occurrence of a word in each document) which will then be used to calculate something termed by TF-IDF, or that's what I understood so far.