Listwise Preference Optimization with Element-wise Confusions for Aspect Sentiment Quad Prediction

2025-11-28

AI Summary

ASQP requires jointly predicting aspect terms, aspect categories, opinion terms, and sentiment polarities as a quadruple, yet existing tagging-based approaches struggle to model complex inter-element dependencies—particularly for higher-order elements like categories and polarities. To address this, we propose an inference-driven generative framework: (1) a unified template augmented with element-specific prefixes to jointly generate the quadruple and natural-language justifications; and (2) listwise preference optimization that leverages syntactically and semantically proximal element-level confounder candidates to enhance structural consistency and relational modeling. Evaluated on four benchmark datasets, our method achieves significant improvements in quadruple accuracy and explanation consistency, while delivering strong interpretability and robustness.

Technology Category

• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.
• Knowledge Representation and Reasoning: Preferences
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Sentiment analysis and opinion mining

Abstract

Aspect sentiment quad prediction (ASQP) is inherently challenging to predict a structured quadruple with four core sentiment elements, including aspect term (a), aspect category (c), opinion term (o), and sentiment polarity (s). Prior methods relying on marker-based prediction struggle with modeling the intricate relationships among elements and experience sharp performance declines when predicting higher-order elements (e.g., c and s) under standard supervised fine-tuning. To address these limitations, we employ reasoning-based generation to output both the quadruple and a natural language rationale under element prefixes within a unified template, encouraging explicit relational reasoning and interpretability. To further enhance element-wise alignment, we introduce a listwise preference optimization framework for improving structural validity and relational coherence. Specifically, we generate element-wise confusable candidates via syntactic and semantic proximity, then train the model with listwise objectives to prefer the gold candidates over closely competing alternatives. Extensive experiments on four benchmark datasets demonstrate that our framework effectively improves quadruple prediction accuracy and explanation consistency.

Problem

• Predicting structured quadruples with four sentiment elements in aspect sentiment analysis
• Addressing performance declines in higher-order element prediction under standard fine-tuning
• Improving structural validity and relational coherence in quadruple prediction accuracy

Innovation

• Reasoning-based generation with element prefixes
• Listwise preference optimization for structural validity
• Element-wise confusable candidates via proximity