A New Paradigm for AIS Data Analysis: Research on Intelligent Methods Based on DeepSeek and GIS

• Published in: 2025 International Conference on Artificial Intelligence and Digital Ethics (ICAIDE) pp. 150 - 154
• Main Authors: Hui, Datao, Miao, Meng, Yao, Gaole
• Format: Conference Proceeding
• Language: English
• Published: IEEE 29.05.2025
• Subjects: Add-in; AIS; Artificial intelligence; Cognition; Costs; Data analysis; Data mining; DeepSeek; GIS; Local Deployment; Python; Software; Software algorithms; Systems architecture; Training
• DOI: 10.1109/ICAIDE65466.2025.11189461

AIS (Automatic Identification System) data serves as the fundamental data for modern maritime activities, and the efficiency and effectiveness of its mining and analysis have become increasingly critical. Currently, the mining and analysis of AIS data suffer from low efficiency and high costs. If only LLM (Large Language Model) is used for data mining and analysis, training a professional domain LLM in advance is required, which incurs substantial time and financial costs. This paper proposes a method for integrating DeepSeek and GIS(Geographic Information System) for data mining and analysis. By leveraging the natural language processing and reasoning capabilities of LLM to generate Python code for invoking GIS algorithm tools, data analysis within GIS software is enabled, which enhances the intelligence level. Meanwhile, it reduces the performance requirements for LLM, saving computing resources and lowering costs.