Автоматизована верифікація знань та виявлення нових знань бізнес-процесів за використанням великих мовних моделей

Abstract

Modern business process management requires systematic knowledge formalization across all maturity levels, yet analytical processing of unstructured textual sources remains largely non-automated. This paper investigates the applicability of large language models for automated extraction and verification of temporal rules from enterprise documentation. The proposed prompt-based approach enables direct integration of discovered rules into a business process knowledge base, reducing dependency on domain experts and accelerating knowledge acquisition during process reengineering. Experimental validation confirms the feasibility of automating both knowledge verification and new knowledge discovery using LLMs.