Discovery Of Association Rules Based On Meteorological Data

Data Mining or Knowledge Discovery emerged in the late 1980s has become a hotspot in the fields of artificial intelligence and database technology. Association rule mining is an important sub-branch of data mining.It has been used widely in selective marketing, telecom business, bank management and so on.Along with the constant development of the cause of meteorological information, a lot of data accumulated in the field of meteorology.How to discovery useful information, which often be hidden in the increasing datas and be overlooked,has become a major task for researchs.There are some relations in meteorological data,which can be mined by different method and under different conditions.This paper briefly introduces the concept and various methods and technique and the development trend of data mining,and then make a basis study of classic relational rule algorithms, mainly focusing on Apriori algorithm. From now on, experts have carried on a great deal of work to improve the efficiency of Apriori. This paper also carries out some improving ideas which have been introduced into the field of meteorology by us. According to the related data of the existing sand dust weather, we use the application relational rule to carry on the mining to the weather data, and have implemented it in the laboratory environment. From the experiment, we get some more meaningful rules, discover the hidden information in the meteorological data, hoping the further research of this field to lay the foundation.In conclusion, only when data mining is used in the process of meteorological data collecting, processing and application to find the rules and models which have some inner relationship, the nature of phenomenon and the mechanism of information can be revealed and recognized. Once effective weather prediction was made and measures against disaster climate were taken instantly by using the rules and models, better service can be provided for scientific decision. From the study, we can see that data mining techniques have great potential in dealing with large amount of meteorological database. Data mining has become an effective means to process meteorological data more effectively.In the future, besides expanding the scope of data mining in the field of meteorology,mining algorithms shoule also be improved.