A Progressive Visual-Logic-Aligned Framework for Ride-Hailing Adjudication

TL;DR

This paper introduces RideJudge, a novel framework for automated ride-hailing dispute adjudication that combines visual reasoning, semantic grounding, and hierarchical decision calibration to improve accuracy and interpretability.

Contribution

The paper presents RideJudge, a comprehensive system integrating semantic grounding, adaptive context optimization, and ordinal-sensitive reinforcement learning for transparent and accurate dispute resolution.

Findings

Achieves 88.41% accuracy, surpassing larger baselines.

Effectively grounds abstract liability concepts into concrete trajectories.

Enhances interpretability and decision calibration in automated adjudication.

Abstract

The efficient adjudication of responsibility disputes is pivotal for maintaining marketplace fairness. However, the exponential surge in ride-hailing volume renders manual review intractable, while conventional automated methods lack the reasoning transparency required for quasi-judicial decisions. Although Multimodal LLMs offer a promising paradigm, they fundamentally struggle to bridge the gap between general visual semantics and rigorous evidentiary protocols, often leading to perceptual hallucinations and logical looseness. To address these systemic misalignments, we introduce RideJudge, a Progressive Visual-Logic-Aligned Framework. Instead of relying on generic pre-training, we bridge the semantic gap via SynTraj, a synthesis engine that grounds abstract liability concepts into concrete trajectory patterns. To resolve the conflict between massive regulation volume and limited context windows, we propose an Adaptive Context Optimization strategy that distills expert knowledge, coupled with a Chain-of-Adjudication mechanism to enforce active evidentiary inquiry. Furthermore, addressing the inadequacy of sparse binary feedback for complex liability assessment, we implement a novel Ordinal-Sensitive Reinforcement Learning mechanism that calibrates decision boundaries against hierarchical severity. Extensive experiments show that our RideJudge-8B achieves 88.41\% accuracy, surpassing 32B-scale baselines and establishing a new standard for interpretable adjudication.