Varma’s team published a paper that exploded the number of choices that could be considered from a search engine from five thousand to ten million. This changed the nature of the game and led to the establishment of a new research area in machine learning called extreme classification. Extreme classification deals with multi-class and multi-label problems involving an extremely large number of choices. Since then, extreme classification has opened a new paradigm for ranking and recommendation applications, such as suggesting related queries on a search engine.