語義匹配在信息監(jiān)管系統(tǒng)中的應(yīng)用研究.pdf

1、華中科技大學(xué)碩士學(xué)位論文語義匹配在信息監(jiān)管系統(tǒng)中的應(yīng)用研究姓名：程莉申請學(xué)位級別：碩士專業(yè)：計(jì)算機(jī)應(yīng)用技術(shù)指導(dǎo)教師：盧正鼎2003.5.9華中科技大學(xué)碩士學(xué)位論文AbstractAneffectivesolutiontoimprovethetraditionalinformationclassifyingsystemperformancesistoclassifyinformationaccordingtothesemanticsort

2、heconceptionsubjectsThemethodofcomputingthesemanticsimilarityamongwordsorsentencesbasedonHownetgivesastrategythattheintegrativesimilarityequalstotheweightedaverageofeverypartsimilarityDecomposinganintegratorintosomeparts

3、，Wecangettwointegrators’similaritybycomputingtheweightedaveragesimilaritiesbetweenthedisassembledcounterpartsInthisway，theintegrativesimilaritybetweentwosemanticexpressionscanbedecomposedintocouplesofprimitivesimilarityP

4、rimitivesimilarityiscomputedbyconvertingthesemanticdistanceThekNearestNeighbor(kNN)textclassificationbasedonLatentSemanticIndexing(LSI)isanapplicationofLSIintheChinesetextclassificationfieldIttakestheadvantagesthatVector

5、SpaceMode(VSM)hasinexpression，andmakesuptheshortageignoringsemanticinVSMThemainideaistopickoutsometeXtkeywords，andanalyzethelatentsemanticrelationsamongthemWecanmatchandclassifythedocumentsbythiswayefficaciouslyThisresea

6、rch，whichisattherequestoftheStateDrugAdministration，canbeappliedtosearchiatricalweb，sitesautomaticallyontheInternetandgiveawarningiftherewereanyillegalinformationThedemosystemhasachievedaverygoodeffectKeyWOrdasemanticcom