Research And Application On Voyage Knowledge Discovery

As the development of world trade and the process of China's socialist modernization economic construction, ocean shipping has become an important national economic sector and has a very important position in the national economy. Task undertaking is the voyage company's major part of their business. The decision of voyage tasks undertaking usually rely on the voyage estimates. Voyage estimate need operational staff have rich experience and knowledge of ocean shipping industry, but it was usually very difficult to have them for primary operational staff. In this view, this thesis wants to discovery the knowledge from the previous voyage instances, to support the voyage estimates.This thesis summarized the predecessors'study of knowledge discovery, and analyse the process of voyage estimates and the contents of voyage instances.According to the requirements of knowledge during voyage estimate and the content of voyage instances, paper find out that the knowledge contain advantage Voyage knowledge, route knowledge, and voyage time knowledge. Then according to the characteristics of the three kinds of knowledge, it selects the appropriate mining method. Advantage Voyage knowledge discovering use decision tree learning methods, analysis the mining object of advantage knowledge from the voyage instance,and the C4.5 and CLS algorithm's application in the discovery of advantage Voyage knowledge; routes knowledge use the method of case-based reasoning, study the case expression, organization, retrieval, and rout case base learning and maintain; voyage time knowledge discovering use statistical clustering method.Study the clustering methods,after the voyage case clustered in accordance with a time-related factors, statistic evey clusters'time as time knowledge. Finally, select the shipping companies'voyage case data, find the voyage knowledge, and use them to solve three relevant issues in estimates to verify the validity of the method.