Automatic Identification System (AIS), as a major data source of navigational data, is
widely used in the application of connected ships for the purpose of implementing maritime situation
awareness and evaluating maritime transportation. Efficiently extracting featured data from AIS
database is always a challenge and time-consuming work for maritime administrators and researchers.
In this paper, a novel approach was proposed to extract massive featured data from the AIS database.
An Evidential Reasoning rule based methodology was proposed to simulate the procedure of
extracting routes from AIS database artificially. First, the frequency distributions of ship dynamic
attributes, such as the mean and variance of Speed over Ground, Course over Ground, are obtained,
respectively, according to the verified AIS data samples. Subsequently, the correlations between
the attributes and belief degrees of the categories are established based on likelihood modeling.
In this case, the attributes were characterized into several pieces of evidence, and the evidence can
be combined with the Evidential Reasoning rule. In addition, the weight coefficients were trained
in a nonlinear optimization model to extract the AIS data more accurately. A real life case study
was conducted at an intersection waterway, Yangtze River, Wuhan, China. The results show that the
proposed methodology is able to extract data very precisely.