From reactive to proactive policy implement method through intelligent enterprise matching

Despite the availability of pro-business policies, many micro, small and medium-sized enterprises (MSMEs) struggle to effectively utilise these resources due to limited capacity. This study proposes an policy-enterprise matching framework that transforms ‘enterprises seeking policies’ approach into a ‘policies seeking enterprises’ mechanism. The mechanism uses large language models (LLMs) as tool for extracting policy restrictive clauses and generating compliance checking strategies. Meanwhile, machine learning methods are used for enterprise risk prediction. The matching framework broadens the research perspective of policy analysis and risk prediction, and provides government staff with decision support based on risk prediction.