基于文本分類的web信息檢索技術(shù)的研究

1、分類號(hào)：UDC：Y1404792TP391密級(jí)：學(xué)校代號(hào)：10150學(xué)號(hào)：20052176峽建交通戈乎碩士學(xué)位論文基于文本分類的WEB信息檢索技術(shù)的研究ResearchonWebInformationRetrievalTechnologyBasedonTextCategorization學(xué)生姓名：導(dǎo)師及職稱：學(xué)科門類：專業(yè)名稱：研究方向：申請(qǐng)學(xué)位級(jí)別：論文答辯日期：學(xué)位授予單位：王軍彭健副教授工學(xué)計(jì)算機(jī)應(yīng)用技術(shù)信息檢索技術(shù)碩士2007年

2、12月23日大連交通大學(xué)大連交通大學(xué)下學(xué)碩十學(xué)位論文AbstractWiththebloomingofInteractinformation，theinformationprocessingisbecominganecessarytoolforpeopletohaveaccesstousefulinformationTraditionalSearchEnginecan’tmeetthedemandthatpeopleincreasing

3、lyneedthecharacteristicinformationsearchserviceTopicorientedtextcategorizationsearchengineemergedwiththetideofthetimesprovidingthemoreprecisesearchserviceinrecentyearsTextiSthemaininformationcarrieronIntemetTextcategoriz

4、ationiSthekeytechniquethattheTopic—orientedsearchengineTextautocategorizationsystemCanorganizeandmanagethetextinformationavailably，locatingtheinformationaccuratelyandrapidly，supportingtheinformationextractingeffectivelyT

5、hepaperresearchesindetailintothetechnologyoftextclassificationbasedonvectorspacemodelTermweighting，termselectionandclassifierconstruction，thekeysoftextclassificationsystem，areintroducedandanalyzedinthepaperThepaperdiscus

6、sesthetraditionalsystemalgorithmoftermweighting：TFIDFtheintroductionoffactorintermweightingispresentedInfeatureselection，thepaperindicatesthereasonthatmutualinformationshowslOWprecisioninclassificationandproposestheimpro

7、vedalgorithmAnewweightadjustmentmethodWasproposedthroughanalysisofTF—IDFBecausetheTF—IDFmethodcouldnotdealwithwordsweightproperly；thenewmethodintroducesfeatureevaluationfunctioninthefeatureweightcomputationandadjuststhef

8、eaturescontributionTheaccuracyofcategorizationwasimprovedusingthenewmethodExperimentalresultsshowthattheimprovedalgorithmsexperimentsoutperformedthetraditionalmethodsinclassificationprecisionTheLucenepackageandopen—sourc