A Novel Lexicon for the Moral Foundation of Liberty

2024-07-16

arXiv.org

AI Summary

Ambiguity in the semantic interpretation of the moral value “freedom” and scarcity of high-quality labeled data lead to biased stance detection in contentious domains such as vaccine hesitancy, climate change, and abortion rights. Method: We construct the first domain-specific lexicon grounded in liberal moral foundations, innovatively integrating word vector similarity with compositional semantic modeling. Our approach leverages over 3,000 manually curated annotations and a multi-source dictionary ensemble to systematically capture cross-platform semantic complexity of freedom-related expressions. Contribution/Results: Extensive experiments demonstrate that our lexicon significantly outperforms mainstream baselines on both in-domain and cross-domain intent identification tasks, markedly improving robustness and generalization. The resource provides a transferable, interpretable semantic foundation for moral reasoning and values-driven stance analysis, advancing computational modeling of moral semantics in social discourse.

Technology Category

• Natural Language Processing: Lexical Semantics and Morphology
• Application Domains: Humanities & Computational Social Science
• Philosophy and Ethics of AI: Morality & Value-based AI

Application Category

• Semantics and Knowledge: Provenance, trust, security and privacy, and ethical issues in managing semantic data
• Web Mining and Content Analysis: Large pretrained models with web data
• Search and Retrieval-Augmented AI: Vertical and domain-specific search

Abstract

The moral value of liberty is a central concept in our inference system when it comes to taking a stance towards controversial social issues such as vaccine hesitancy, climate change, or the right to abortion. Here, we propose a novel Liberty lexicon evaluated on more than 3,000 manually annotated data both in in- and out-of-domain scenarios. As a result of this evaluation, we produce a combined lexicon that constitutes the main outcome of this work. This final lexicon incorporates information from an ensemble of lexicons that have been generated using word embedding similarity (WE) and compositional semantics (CS). Our key contributions include enriching the liberty annotations, developing a robust liberty lexicon for broader application, and revealing the complexity of expressions related to liberty across different platforms. Through the evaluation, we show that the difficulty of the task calls for designing approaches that combine knowledge, in an effort of improving the representations of learning systems.

Problem

• Develops a lexicon for liberty moral foundation analysis
• Evaluates lexicon on manually annotated in-domain data
• Combines word embedding and compositional semantics approaches

Innovation

• Novel Liberty lexicon with manual annotations
• Combined lexicon using embedding and semantics
• Enriched annotations for broader application scenarios