Faithful by Definition: Emotion Analysis via Natural Semantic Metalanguage Explications

2026-07-01

AI Summary

This work proposes an interpretable emotion analysis framework grounded in Natural Semantic Metalanguage (NSM), addressing the limitation of existing classifiers whose explanations are typically post-hoc and fail to reflect genuine decision logic. The approach employs a structured parser to map input text into a twelve-slot semantic script, followed by causal and verifiable emotion classification based on a fixed rule set derived directly from emotion semantic definitions. This method establishes, for the first time in emotion analysis, a faithful explanation mechanism that is both causally transparent and definitionally grounded, ensuring full auditability of the decision process. Experimental results show that a parser fine-tuned on crowdsourced event descriptions achieves 0.33 overall accuracy and 0.48 selective accuracy, demonstrating its capacity to deliver reliable interpretability without sacrificing reasonable performance. The study also introduces EmoExpl-1200, a dataset enriched with validation metadata.

Abstract

Explanations for emotion classifiers are usually produced post hoc, with no guarantee that they reflect the computation behind the label. We present an explication interface for event-based emotion analysis. A parser maps the input text to an explication, a short script in the closed vocabulary of Natural Semantic Metalanguage organized into twelve typed slots, and a fixed decision list of rules transcribed from published semantic definitions computes the label from the explication alone. The faithfulness guarantee is therefore causal and definitional, while all empirical risk lives in the learned parser, which the per-line entailment interface makes auditable against the input. On crowd-sourced event descriptions, our fine-tuned parser reaches 0.33 accuracy and 0.48 selective accuracy on a small held-out set, suggesting that the interface trades insignificant accuracy difference to a black-box model for a verifiable, inspectable decision basis for first-person event-based emotion analysis. We also release EmoExpl-1200 with per-line verification metadata and the full rule set.

Problem

• emotion analysis
• faithful explanation
• Natural Semantic Metalanguage
• explication
• classifier interpretability

Innovation

• Natural Semantic Metalanguage
• faithful explanation
• explication interface
• rule-based emotion analysis
• auditable NLP