Learning-to-Rank

A machine learning approach that directly optimizes the quality of ranking results by training a model to learn the relative order of items. It is commonly used in the ranking stage of search and recommendation systems.