RESEARCH ON THE CONSTRUCTION OF A KNOWLEDGE GRAPH FOR DECISION-MAKING ON COMPLIANCE WITH RESTRICTIVE MEASURES AND COUNTERMEASURES AGAINST ILLEGAL SANCTIONS IN THE RUSSIAN NATURAL GAS INDUSTRY BASED ON THE DEEPSEEK AI MODEL

This article examines the pressure of Western restrictive measures and countermeasures affecting the Russian gas industry in the context of the Russian-Ukrainian conflict. It proposes a composite decision-support system that integrates a large language model (DeepSeek AI), knowledge graph technology, and a compliance reasoning algorithm. The construction of a compliance knowledge graph, encompassing sanction rules, corporate entities, and energy trade routes and augmented by a graph neural network mechanism, enables the intelligent identification of the impact of illegal sanction policies and the planning of responsive strategies. This study presents an "explainable and reasonable" framework of a sanctions and counter-sanctions graph that Chinese energy companies can use to anticipate the sanction-related risks they may encounter when cooperating with Russia, the Middle East, and Central Asia. At the same time, it introduces a compliance decision-making tool that deeply integrates language understanding capabilities of large models with a graph reasoning system, providing a cross-integration path for disciplines such as law, artificial intelligence, international relations, and energy economics in both China and Russia.