A Comparison of Contextual and Non-Contextual Preference Ranking for Set Addition Problems

TL;DR

This paper compares contextual and non-contextual preference ranking methods for set addition problems, demonstrating that a triplet network approach outperforms a twin network in modeling human preferences in a card game.

Contribution

The paper introduces and empirically compares two Siamese network architectures for preference ranking in set addition, highlighting the superiority of the triplet network.

Findings

Triplet network outperforms twin network in preference ranking.

Both models outperform previous methods on the task.

Contextual modeling improves preference prediction accuracy.

Abstract

In this paper, we study the problem of evaluating the addition of elements to a set. This problem is difficult, because it can, in the general case, not be reduced to unconditional preferences between the choices. Therefore, we model preferences based on the context of the decision. We discuss and compare two different Siamese network architectures for this task: a twin network that compares the two sets resulting after the addition, and a triplet network that models the contribution of each candidate to the existing set. We evaluate the two settings on a real-world task; learning human card preferences for deck building in the collectible card game Magic: The Gathering. We show that the triplet approach achieves a better result than the twin network and that both outperform previous results on this task.