# Preference Modeling with Context-Dependent Salient Features

arXiv: https://arxiv.org/abs/2002.09615 · HTML: https://mlsanity.com/paper/2002.09615/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 22 Feb 2020.

## Built on by (most cited first, 8 of 8)

- [Graph-Based Methods for Discrete Choice](https://mlsanity.com/paper/2205.11365/) (2022)
- [Transformer Choice Net: A Transformer Neural Network for Choice Prediction](https://mlsanity.com/paper/2310.08716/) (2023)
- [Leveraging Sparsity for Sample-Efficient Preference Learning: A Theoretical Perspective](https://mlsanity.com/paper/2501.18282/) (2025)
- [What Does Preference Learning Recover from Pairwise Comparison Data?](https://mlsanity.com/paper/2602.10286/) (2026)
- [Choice Set Confounding in Discrete Choice](https://mlsanity.com/paper/2105.07959/) (2021)
- [DeepHalo: A Neural Choice Model with Controllable Context Effects](https://mlsanity.com/paper/2601.04616/) (2026)
- [A Theory of Tournament Representations](https://mlsanity.com/paper/2110.05188/) (2021)
- [Similar Models, Broken Odds: Where Luce's Axiom Fails in Chatbot Arena](https://mlsanity.com/paper/2610.19095/) (2026)

## Abstract (arXiv)

We consider the problem of estimating a ranking on a set of items from noisy pairwise comparisons given item features. We address the fact that pairwise comparison data often reflects irrational choice, e.g. intransitivity. Our key observation is that two items compared in isolation from other items may be compared based on only a salient subset of features. Formalizing this framework, we propose the salient feature preference model and prove a finite sample complexity result for learning the parameters of our model and the underlying ranking with maximum likelihood estimation. We also provide empirical results that support our theoretical bounds and illustrate how our model explains systematic intransitivity. Finally we demonstrate strong performance of maximum likelihood estimation of our model on both synthetic data and two real data sets: the UT Zappos50K data set and comparison data about the compactness of legislative districts in the US.