ABSTRACT: In recent years, with the development of information technology, businesses have accumulated a lot of useful historical data, as the shipping industry does. These data can be found deposited a large number of "knowledge", for example, Shipping records for historical information, Ship-Port relations information, Ship-ship relations information, Port & shipping route relations, Shipping route information. It can provide intellectual support to shipping informatization development. Association rules in data mining technology is one of important technologies. The technology, based on statistical methods, can mine the associated and implied "knowledge" from data warehouse ,which has a large number of accumulated data. Apart from this, the technology can also play an important role in the prediction. In this paper, based on FP-growth algorithm, we improve it forming Relevent ships routes. From the prevalent perspective of data mining, deal with the corresponding vessels' dynamic information, acquired from the AIS, such as data collection, data statistics. On this basis, get the ship-port relation and ship-ship relation after a certain level of data analysis, processing, handing. Furthermore, this paper use the numerous historical ship-port relation and ship-ship relation to build a mathematical model on the ship-port and ship-ship relation. And use the improved association algorithm, FP-growth algorithm, to acquire the strong association rules between ship-port and ship-ship, and eventually mine the similarity of the ship route. Main points of this paper as follows: Collect ,count and check the data, which is from ship dynamic information; Establish the mathematical model between ship-port and ship-ship relation; Improve the algorithm; Analyse the similarity of ship route more accurately using the improved algorithm.

KEYWORDS: Safety at Sea, Marine Navigation, Automatic Identification System (AIS), Mathematical Model, Collision Avoidance, Route Analysis, Statistical Method, Algorithms

REFERENCES