Research on Key Technologies of Data Mining System Based on Web Service

As more and more data accumulated, companies are eager to dig out the knowledge behind a large amount of data to support decision-making. The existing data mining tools such as IBM's Intelligence Miner and SAS Enterprise Miner, although they provide a comparatively rich mining function, these tools cannot dig out distributed and highly heterogeneous data on Internet/Extranet, and it is not effectively connected with the operating system integration and target is not strong. If enterprises use these tools, the invest is heavy, some of the mining work is not used, and the algorithm library is very difficult to be upgraded. This paper presents a data mining system framework based on Web services. It can be well integrated with the original operating system, and it can excavate the data in the distributed database. Additionally, it has the advantages of cross platform, cross language, easy deployment and dynamic management algorithm library.