Research On Application Of Data Mining Technology In Enterprise

The data mining technology is an interdisciplinary studies.Which is rapidly develop in recent years, involves thedatabase, statistics, artificial intelligence and the machinelearning and so on many domains. The applied universality of thecomputer produced a great deal of data, The data mining is thetechnique which makes use of above-mentioned science to carry on thegreat data quantity processing. The data mining application isextremely broad, cheats from the credit card to the taxaffairsexamines, Conghua studies the molecular structure therecognitionteam to be on site to replace the team member, from theagriculturalproduction forecast to the gene classification, The data miningtechnology will play more and more main role to future society's eachdomain.Our country data mining technology on the one hand is thescientific research institution pauses in the scholarly research,on the otherhand is the use overseas company's software productsolution concrete questions. In order to raise the academic level,the scientific researchers have to carry on the high level but thevery difficult practical algorithm research, enterprise practicaloverseas company's software issteadiest. But the data miningtechnology needs in the solution actual problem process is themature technology adds on in view of the concrete question revision,therefore the domestic urgent need logarithm carries on theanalysis according to the excavation concretetechnology, mainlywas Fudan University, the Beijing University, the Harbin industrialuniversity, northeast University, the southeast university and soon some development facilities already aimed at the data stream eachkind of related question to launch more thorough research, But fewhave the technique result of international influence.Data mining raise our application of date from the low levelsimple inquiry, promotes to excavate the knowledge from the data,provides the policy-making support service. The data mining processmay divide into: Question stage and so on definition, datapretreatment, data mining as well as result explanation andappraisal, at present the logarithm mainly concentrates according to the excavation research to the excavation technology, the excavationalgorithm, the excavation language and so on, but the data miningto the data which processes has the strict quality requirement infact, the data pretreats very important in the data miningprocess. According to counts, in a complete data mining process, thedata pretreatment must spend 60% about the time, but the afterexcavation work only occupies the total work load about 10%, the datapretreatment mainly will include the data cleaning, the integrationand the selling and buying of real esgate within the same family, the data cleaning is processes in the data the omission and cleansthe dirty data, the data integration carries on the multi- data poolsin data merge processing, solution semantics fuzziness and theentiresynthesis consistent data storage, the data selling and buyingof realesgate within the same family distinguishes the dataacquisition whichwill need to excavate, will reduce the processingscope.The oil field computer network construction is consummating dayby day, Oil-field development and produce a great deal of informationwithin research work, it has already been delivered by internet, thequantity of the information sharing and the quality enhanced greatly, developing the multi- servers based on the great information contentdatabase information network, provided the multi-positionsmonitoring informationconstruction quality and the managementprocess method, promptly comprehensively accurately understoodgrasped the daily user to thedatabase service condition, caused inthe on-line server the datamessage management to change visible andtransparent, through thecomputer program movement substitutiontradition artificialinformation management pattern, was thequestion which was urgentneeds to solve.This thesis mainly studied several following contentses of:(1) Through the research on the domestic and foreign data streamstechnology system, summarizes goes beyond the great wall to the datamining as well as the data management system (DSMS) the essentialfeature, the model as well as the essential technology and so on.(2) Daqing tests the explanation appraisal center eachproduction department use the PDPMIS system deliver a great deal ofdate to the center continualy everyday, the present paper considersas the mobile data set a special data stream, (DSMS) makes the data management system for the traditional database systemexpansion. Proposed based on the data stream excavation oil fieldnetwork monitorand control nmac system model, the model increasedthe use rconnection, the data monitoring, inquired system module andso on the processor, memory block, in the data stream excavationmodule through the Operators connection, the system could throughfilter, gather, the classification, or neglected certain specificdata elements.(3)The oil-field produce medium, more interested in problem isrecently which modes reflect more and unusually, this thesis studiedthe withdraw of mode of the data abnormality. The area value conditionof enactment is an experience value of oil extraction factory severalyears and adopt XML to combine the way to build up that the data flowssystem system of can expand sex. Introduced one piece in the meantimeaccording to scoop out environment-YaLE(the LearningEnvironment ofthe Yet Another) and aim at the sequence data of time a processingsolid hour and imitate of the Conceptdrift plugin of a plug-in toolof concept drift. Combined to the machine KBS-data of an integrationstudy's flow(the for Streams of the Bayesian Boosting) experiencekind example and one piece can adapt according to flow of time windowway allocation document.(4)Introduced design and realization of the supervision systemof the oil-field network, the system will report to the police thearea value of condition enactment to keep a XML document and adoptthe DOM interview XML document; Made use of ORACLE audit and triggera machine, JOB etc. function to combine JAVA procedure to carry onoverall supervision to the circumstance, production dataquality...etc, of the database operation in the network, and carryon reporting to the police a hint to the excrescent circumstance, in order to give attention to both time and velocity and actual need, combine solid hour with postpone of way, adopt a layering to reportto the police. Take actual item as to rely on, the management systemof the convection data with flowed data excavation a research toprovide some experience and thinking.