基于ATN算法和潛在語義索引的不良信息過濾系統(tǒng)研究.pdf

1、西南交通大學碩士學位論文基于ATN算法和潛在語義索引的不良信息過濾系統(tǒng)研究姓名：蘇再強申請學位級別：碩士專業(yè)：計算機應用技術指導教師：朱焱201105西南交通大學碩士研究生學位論文第1I頁AbstractInrecentyears，theInternethasmaderapiddevelopment，informationintheIntemetisincreasinglyabundant，growingexponentiallyunt

2、ilnowtherearehundredsofmillionsofwebsitesaroundtheworldHowevereverythinghastwosides，whiletherapiddevelopmentoftheIntemet，alotofbadinformationunbridledspreadtOthetheInternet，hasbroughtabouttremendousnegativeimpacttOthesoc

3、ialorderandpeople’SnormallifeThispaperdescribedthegrimsituationofthecurrentIntemetSummeduptheconcepts，characteristics，andrelatedproductsofthecurrentthreecommonobjectionableinformationfilteringtechnologyAnalyzedandcompare

4、dthefilterresults，advantagesanddisadvantagesofthethreetypesoffilteringtechnologyOnthisbasis，wedesignedanobjectionableinformationfilteringsystembasedOilAugmentedTransitionNetwork(ATN)andLatentSemanticIndexing(LSI)TheATNal

5、gorithmusedinthesysteminsteadofthetraditionalsegmentationbasedonstringmatchingalgorithms，thisalgorithmisbasedonChineselexical，grammaticalrules，SOtheresultsofsegmentationhasahigheraccuracyReplacingthesimpletraditionalvect

6、orspacemodeltOfilterinformationthealgorithmhasagoodeffectondealingwithChinese”Polysemy”and”Morewordonemeaning”Inordertoputtheresultsoffilterintopractice，thesystemmarkstheresultsbyPICSlabelandstoresitinthedatabaseforappli

7、cationFinallytoverifythefilterperformanceofthesystemwedidexperimentsonDataSetsofChineseEMails(CDSCE)providedbyChinaEducationandResearchNetworkEmergencyTeam(CCERT)，theexperimentresultsshowthatthesystemdesignreasonableandf