HORAE: A Domain-Agnostic Language for Automated Service Regulation

TL;DR

Horae introduces a domain-agnostic regulation language and a formal semantics framework to model cross-domain service rules, enabling automated reasoning and violation detection. The approach combines RuleGPT, a triplet of fine-tuned models via LoRA, to translate natural-language regulations into Horae by extracting basic events, logical relations, and syntactic patterns, all validated through SMT-based consistency checks. Experimental results on the SRR-Eval benchmark show that RuleGPT with 7B parameters can outperform GPT-3.5 and match GPT-4o on key tasks, demonstrating feasibility for end-to-end automated intelligent service regulation. The work provides open-source datasets and tools, paving the way for scalable, domain-unified regulation modeling and automated enforcement across modalities, including multimodal rules.

Abstract

Artificial intelligence is rapidly encroaching on the field of service regulation. However, existing AI-based regulation techniques are often tailored to specific application domains and thus are difficult to generalize in an automated manner. This paper presents Horae, a unified specification language for modeling (multimodal) regulation rules across a diverse set of domains. We showcase how Horae facilitates an intelligent service regulation pipeline by further exploiting a fine-tuned large language model named RuleGPT that automates the Horae modeling process, thereby yielding an end-to-end framework for fully automated intelligent service regulation. The feasibility and effectiveness of our framework are demonstrated over a benchmark of various real-world regulation domains. In particular, we show that our open-sourced, fine-tuned RuleGPT with 7B parameters suffices to outperform GPT-3.5 and perform on par with GPT-4o.

Figures (7)

• Figure 1: Conventional plug-and-play methods are often confined to distinct models for specific domains, thus requiring extensive retraining and resource expenditure. In contrast, Horae acts as a unified specification language to model regulation rules in a domain-agnostic fashion.
• Figure 2: Horae-steered intelligent service regulation.
• Figure 3: The overall process of automated transformation using the fined-tuned RuleGPT.
• Figure 4: Visualization of data in \ref{['tab:experiment-results-event']} (Q-Ins abbreviates Qwen2.5-7B-Ins). Every scattered boxplot depicts the corresponding column of \ref{['tab:experiment-results-event']} with its five-number summary.
• Figure 5: An illustration of the video-to-text pipeline.

Theorems & Definitions (1)

• Remark 1