Data Reduction And The Application In Auditing Of The Social Security Network

Nowadays the world has already entered the Internet information age. The high-speed developments of the computer and Internet technologies have made data and information in every fields increase rapidly. And the uncertainty of data and information systems is more obvious because of people's participation. How to mine the underlying and valuable information from massive, disorganized and strong jamming data gives an unprecedented challenge in the human intelligence information processing field.In the various aspects of research on the rough set theory, the reduction of the information system and the minimum reduction are not only the basic problems, but also the most pivotal and intractable problems. The majority of traditional rough set methods are not unified with the database, so it causes the rough set theory hard to fit the practical size of database, therefore the study of the efficient algorithms for attribute reduction based on database is one of the practical goals in this thesis.This thesis, which introduces attribute selection and rough set, analyzes feature selection algorithms based on rough set model. Because of the traditional rough set model not combining with the Relational Database Systems, all the calculation of basic operations are carried out in files without the cost of high-performance database assembly operation. But Rough Set Model based on database systems redefined nuclear attributes and reduction on the basis of the relational algebra. So efficient operator set oriented Database System is used. Efficient Operator Set Oriented Database System is used to avoid redefined nuclear attributes and reduction on the relation algebra basis of Rough Set Model based on database systems. Based on the rough set model database system, the heuristic algorithm based on database system is improved and inspiration Information of the attribute importance is changed from static to dynamic, which can effectively achieve the smallest reduction of attributes. Meanwhile, a rough system DBReduced intensive briefing based on database system was developed with Jbuilder and Java. Then the system was tested. In addition, this system has been applied to the social security network audits and achieved the expect results.