Learning the Value Systems of Societies from Preferences

2025-07-28

AI Summary

Conventional approaches to societal value modeling aggregate individual values naively, ignoring inherent heterogeneity across population subgroups. Method: We propose a heuristic deep clustering framework that jointly learns preference-based representations and discovers latent group structures, modeling societal values as a collection of interrelated yet distinct multi-group value systems—enabling end-to-end inference from individual decision data to both shared value foundations and group-specific value configurations. Contribution/Results: Evaluated on real-world urban mobility decision data, our method effectively disentangles universal value consensus from group-specific value orientations, yielding interpretable, representative, and structurally grounded societal value models. It significantly improves model transparency and fidelity compared to baseline aggregation methods, offering a principled alternative for value-aware behavioral modeling in complex social systems.

Technology Category

• Humans and AI: Learning Human Values and Preferences
• Knowledge Representation and Reasoning: Preferences
• Machine Learning: Learning Preferences or Rankings

Application Category

• User Modeling, Personalization and Recommendation: Accountability, Transparency, and Ethics for personalization
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: The sharing economy

Abstract

Aligning AI systems with human values and the value-based preferences of various stakeholders (their value systems) is key in ethical AI. In value-aware AI systems, decision-making draws upon explicit computational representations of individual values (groundings) and their aggregation into value systems. As these are notoriously difficult to elicit and calibrate manually, value learning approaches aim to automatically derive computational models of an agent's values and value system from demonstrations of human behaviour. Nonetheless, social science and humanities literature suggest that it is more adequate to conceive the value system of a society as a set of value systems of different groups, rather than as the simple aggregation of individual value systems. Accordingly, here we formalize the problem of learning the value systems of societies and propose a method to address it based on heuristic deep clustering. The method learns socially shared value groundings and a set of diverse value systems representing a given society by observing qualitative value-based preferences from a sample of agents. We evaluate the proposal in a use case with real data about travelling decisions.

Problem

• Aligning AI systems with diverse human value systems
• Learning societal value systems from group preferences
• Automating value system modeling via deep clustering

Innovation

• Learning societal value systems from preferences
• Heuristic deep clustering for value system diversity
• Shared value groundings via preference observation