FOLD-TR: A Scalable and Efficient Inductive Learning Algorithm for Learning To Rank
This work addresses ranking tasks for domains requiring interpretability and mixed data handling, but it appears incremental as it builds upon the existing FOLD-R++ algorithm.
The authors tackled the problem of learning to rank with mixed-type data by introducing FOLD-TR, a scalable and efficient inductive learning algorithm that generates explainable normal logic programs, resulting in a method that directly handles numerical and categorical data and provides native justifications for item comparisons.
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.