Chinese-Russian bilingual LLM for the China-Russia pipeline and its application in disaster scenarios
Year: 2025
Type: conference-paper
Abstract: The China-Russia Pipeline is a significant economic and geopolitical infrastructure project that is essential to both nations' energy security and the advancement of bilateral relations. The development of a new Chinese-Russian bilingual LLM that focuses on catastrophe prevention, response, and recovery on this crucial pipeline is described in this research. To do this, we build a top-notch Chinese-Russian bilingual knowledge base to improve the LLM. To improve the model's reasoning and decision-making abilities, we also employ knowledge graphs and idea chain reasoning approaches. Given these considerations, the goal of this work is to provide a practical MMP of an LLM that is backed by empirical evaluations demonstrating that our approach outperforms general-purpose LLMs in terms of decision-making capabilities and even surpasses them in answering pipeline-related domain-specific queries.
Authors: Haoran Li, Duo Shang, Jun Luo, Chunyu Lu, Huihong He, Xin Hui, Tianran Chen, Ruhui Shi
Language: English
Topic: Topic Modeling
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4410812635