**Preface to "Data/Knowledge-Driven Behaviour Analysis for Maritime Autonomous Surface Ships"**

With the development of artificial intelligence and ICT, ships are expected to be smarter than traditional ships in the future, as they can autonomously navigate from one point to another point in the waters. Several studies have developed various systems to achieve such goals. As we can see, the MASS can handle many tasks with explicit references, such as speed following, course keeping, path following, etc. However, the MASS does have some limitations in operating some complicated tasks that need the machine to make decisions and adjust its reference according to the recognized traffic scene, such as collision avoidance, emergent operations, etc. In this process, we found that the recognition and prediction of ship behavior are essential for the recognition of traffic scenes, which will influence the decision outcomes. For instance, when two ships encounter each other, the give-way ship's behavior will influence the decision of the stand-on ships. Thus, we believe the study on ship behavior would benefit the development of MASS.

Recently, the developments of equipment onboard ships enrich our data source to analyze ship behavior, such as radar, Automatic Identification Systems (AIS), CCTV, etc. These Maritime traffic data (e.g., radar data, AIS data, CCTV data) provide designers, officers on watch (OOW), and traffic operators with extensive information about the states of ships at present and in history, which are a treasure for behavior analysis. Additionally, the development of knowledge analysis tools, e.g., Fuzzy systems, knowledge graphs, etc., offer a new insight to analyze the ship's behavior based on human knowledge, e.g., navigation rules and regulations. Combining multisource heterogeneous big data and artificial intelligence techniques inspires innovative and important means for understanding ship behavior and developing MASS. Thus, under the support of the Key R&D Program of Zhejiang Province (China) through Grant No. 2021C01010, this reprint collects 12 papers working on data/knowledge-driven behavior analysis for MASS and its applications, including data-driven behavior modeling, knowledge-driven behavior modeling, multisource heterogeneous traffic data fusion, risk analysis and management of MASS, etc.
