Integrating Knowledge Graphs for Enhanced Drilling Accident Retrieval and Risk Prediction in China's Oil and Gas Industry

This paper examines the transformative journey of China's oil and gas sector, focusing on the challenges associated with its growth into a major producer and the world's largest importer due to expansive industrial demands. We delve into the intricacies of drilling operations and the frequency of anomalies that lead to accidents, citing significant incidents and highlighting the reliance on experiential judgments in risk assessment. The thesis proposes the adoption of a knowledge graph as a structured semantic knowledge base to process and integrate data, aiming to improve the retrieval, visualization, and analysis of accident information. Through the construction of an ontology for the drilling accident domain using Protégé and the development of an intelligent question-answering system employing Neo4j, this research offers a novel approach to drilling accident retrieval and risk reasoning. The system's capability to predict potential hazards based on abnormal signs is demonstrated, underscoring its practical value in proactive risk management for drilling operations.