Incorporating Legal Logic into Deep Learning: An Intelligent Approach to Probation Prediction

2025-08-17

AI Summary

Existing intelligent judicial assistance systems lack task-specific methods for probation prediction and often neglect legal reasoning, relying excessively on purely data-driven modeling. To address this, we propose MT-DT—a novel multi-task deep learning framework that uniquely integrates formal legal logic with the dual-track theory of punishment. We construct a manually annotated dataset covering both factual descriptions of offenses and statutory probation eligibility criteria, enabling joint modeling of legal reasoning processes and textual semantic features. MT-DT jointly optimizes two tasks: probation eligibility classification and identification of legally salient elements. It significantly outperforms baseline models across multiple evaluation metrics. Moreover, legal consistency analysis confirms its decision interpretability and jurisprudential soundness. This work establishes a new paradigm for legal AI—logically embeddable, empirically verifiable, and grounded in domain-specific legal principles.

Technology Category

• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Reasoning under Uncertainty: Relational Probabilistic Models
• Machine Learning: Statistical Relational/Logic Learning

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Probation is a crucial institution in modern criminal law, embodying the principles of fairness and justice while contributing to the harmonious development of society. Despite its importance, the current Intelligent Judicial Assistant System (IJAS) lacks dedicated methods for probation prediction, and research on the underlying factors influencing probation eligibility remains limited. In addition, probation eligibility requires a comprehensive analysis of both criminal circumstances and remorse. Much of the existing research in IJAS relies primarily on data-driven methodologies, which often overlooks the legal logic underpinning judicial decision-making. To address this gap, we propose a novel approach that integrates legal logic into deep learning models for probation prediction, implemented in three distinct stages. First, we construct a specialized probation dataset that includes fact descriptions and probation legal elements (PLEs). Second, we design a distinct probation prediction model named the Multi-Task Dual-Theory Probation Prediction Model (MT-DT), which is grounded in the legal logic of probation and the extit{Dual-Track Theory of Punishment}. Finally, our experiments on the probation dataset demonstrate that the MT-DT model outperforms baseline models, and an analysis of the underlying legal logic further validates the effectiveness of the proposed approach.

Problem

• Lack of dedicated probation prediction methods in IJAS
• Limited research on factors influencing probation eligibility
• Data-driven approaches overlook legal logic in judicial decisions

Innovation

• Constructs specialized probation dataset with legal elements
• Designs MT-DT model based on Dual-Track Theory
• Integrates legal logic into deep learning for prediction