Human-In-The-Loop Learning of Qualitative Preference Models
This addresses the need for interpretable preference modeling in combinatorial domains, but it appears incremental as it builds on existing models like lexicographic preference trees.
The paper tackles the problem of helping users understand decision-making processes by developing a human-in-the-loop framework for learning qualitative preference models from user behavioral data, with results including interactive visualization of explainable graphic models like lexicographic preference trees.
In this work, we present a novel human-in-the-loop framework to help the human user understand the decision making process that involves choosing preferred options. We focus on qualitative preference models over alternatives from combinatorial domains. This framework is interactive: the user provides her behavioral data to the framework, and the framework explains the learned model to the user. It is iterative: the framework collects feedback on the learned model from the user and tries to improve it accordingly till the user terminates the iteration. In order to communicate the learned preference model to the user, we develop visualization of intuitive and explainable graphic models, such as lexicographic preference trees and forests, and conditional preference networks. To this end, we discuss key aspects of our framework for lexicographic preference models.