The Port Production Data Analysis System Based On Datamining

At present,the competition between ports is intensifying,and the data analysis are high requirement.How to effectively mine the hidden rules of port data is a urgent subject.In this paper,the analysis system of port data is designed based on data mining.It can effectively mine the association and scientifically make decision-making of port data.Firstly,the data warehouse of port data is designed,and the conceptual model,logical model and physical model are designed.Secondly,the clustering is applied of port data.The data screening,data cleaning and data conversion are implemented based on the original port data.The data feature extraction and data preprocessing are completed.The data of port is classified by using k-means.The k data with the greatest entropy are selects as the clustering center in this algorithm,and uses the average contour coefficient to determine the clustering center.In each cluster,the similarity of port data is great.So the managers can better understand the characteristics of each port business.Again,the association relations are mined by using association rule analysis.The thought of first divide and integrate later is introduced.The frequent patterns of port data are mined by Apriori algorithm.To find out the trend of port business,the hidden relationship between port data are applied.Finally,based on the above analysis,data mining technology is used to design the port production data analysis system.