📄 中文摘要
该研究提出了一种上下文感知的轨迹抽象框架，旨在将从自动识别系统（AIS）收集的原始船舶轨迹数据转化为结构化且语义丰富的表示形式。这种表示形式不仅便于人类理解，还能直接供机器推理系统使用。框架将嘈杂的AIS序列分割为不同的航程，每个航程由干净的、带有移动性注释的片段组成。每个片段进一步融入多源上下文信息，包括附近的地理实体、离岸导航特征和天气条件。这种表示方式能够支持使用大型语言模型（LLMs）生成受控的自然语言描述。研究还通过多种LLM对生成描述的质量进行了实证检验。
上下文丰富的船舶轨迹自然语言描述
出处: Context-Enriched Natural Language Descriptions of Vessel Trajectories
发布: 2026年3月16日
📄 English Summary
Context-Enriched Natural Language Descriptions of Vessel Trajectories
This study proposes a context-aware trajectory abstraction framework aimed at transforming raw vessel trajectory data collected from the Automatic Identification System (AIS) into structured and semantically enriched representations. These representations are interpretable by humans and directly usable by machine reasoning systems. The framework segments noisy AIS sequences into distinct trips, each consisting of clean, mobility-annotated episodes. Each episode is further enriched with multi-source contextual information, including nearby geographic entities, offshore navigation features, and weather conditions. Such representations support the generation of controlled natural language descriptions using large language models (LLMs). The quality of these generated descriptions is empirically examined using several LLMs.