FOLD-TR: A Scalable and Efficient Inductive Learning Algorithm for Learning To Rank

TL;DR

FOLD-TR is a scalable, explainable inductive learning algorithm designed for ranking tasks, capable of handling mixed data types and providing justifications for item comparisons.

Contribution

It extends FOLD-R++ to a ranking framework, enabling explainable, efficient learning for ranking with mixed numerical and categorical data.

Findings

Handles mixed data types directly

Provides native explanations for rankings

Scalable and efficient for large datasets

Abstract

FOLD-R++ is a new inductive learning algorithm for binary classification tasks. It generates an (explainable) normal logic program for mixed type (numerical and categorical) data. We present a customized FOLD-R++ algorithm with the ranking framework, called FOLD-TR, that aims to rank new items following the ranking pattern in the training data. Like FOLD-R++, the FOLD-TR algorithm is able to handle mixed-type data directly and provide native justification to explain the comparison between a pair of items.