Missing-by-Design: Certifiable Modality Deletion for Revocable Multimodal Sentiment Analysis

2026-02-17

AI Summary

This work addresses the challenge of removing specific modality data in multimodal sentiment analysis to comply with privacy regulations. To this end, the authors propose a unified framework that enables verifiable modality forgetting without requiring full model retraining. The approach integrates attribute-aware embeddings, generative reconstruction, and saliency-driven parameter updates to perform precise, “surgical” modifications of model parameters, effectively erasing targeted modality information while preserving overall functionality. Experimental results on standard benchmarks demonstrate that the framework maintains strong predictive performance even under modality removal, successfully balancing privacy preservation with model utility. This practical trade-off offers a viable solution for deploying privacy-compliant multimodal systems without significant degradation in effectiveness.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Security and Privacy: Data transparency and provenance

Abstract

As multimodal systems increasingly process sensitive personal data, the ability to selectively revoke specific data modalities has become a critical requirement for privacy compliance and user autonomy. We present Missing-by-Design (MBD), a unified framework for revocable multimodal sentiment analysis that combines structured representation learning with a certifiable parameter-modification pipeline. Revocability is critical in privacy-sensitive applications where users or regulators may request removal of modality-specific information. MBD learns property-aware embeddings and employs generator-based reconstruction to recover missing channels while preserving task-relevant signals. For deletion requests, the framework applies saliency-driven candidate selection and a calibrated Gaussian update to produce a machine-verifiable Modality Deletion Certificate. Experiments on benchmark datasets show that MBD achieves strong predictive performance under incomplete inputs and delivers a practical privacy-utility trade-off, positioning surgical unlearning as an efficient alternative to full retraining.

Problem

• revocable multimodal learning
• modality deletion
• privacy compliance
• user autonomy
• multimodal sentiment analysis

Innovation

• revocable multimodal learning
• certifiable deletion
• surgical unlearning
• modality reconstruction
• privacy-preserving AI