私有信息检索算法分析-analysis of private information retrieval algorithm.docxVIP

AbstractWiththeextensiveapplicationofinformationtechnology,publiclyaccessibledatabasesandsearchengineersareindispensableresourcesforretrievingthelatestinformation.However,becauseoftheshortcomingsoftraditionalprivateinformationretrievalmodel,itisdifficulttoapplytolargedatabasesandsearchengines.Therefore,theresearchonnewandpracticalprivateinformationretrievalmodelandalgorithmhasimportantsignificance.Accordingtotheanalysisoftheimplementedprivateinformationretrievalsystemsandfunctionalrequirementsofwordsemanticsimilarity-basedprivateinformationretrievalsystem,weproposeanewwordsemanticsimilarity-basedprivateinformationretrievalmodel.Thenwetakerelatedanalysistothepartsofwordsemanticsimilaritycomputing,boguskeywordselectionstrategy,queryinformationhiding,andthequeryresultsfilteringandgivetheoverallarchitectureofthesystem.Thearchitectureofthesystemincludeswordsemanticsimilaritycomputingmodular,queryprocessingmodularandpagecrawlingandfilteringmodular.WeintroducetheimplementofwordsemanticsimilaritycomputingalgorithmbasedonWordNetandHowNet.OnthebasisoftheexistingwordsemanticsimilaritycomputingalgorithmbasedonWordNet,wetaketheinfluencingfactorofthedepthofthenodeintoaccount.ThenweapplythisalgorithmtotheprimitivesimilaritycomputingofHowNet.Theexperimentsshowthatimprovedalgorithmgetsmoreaccuratesimilarityresults,accordingmorewiththehumanbeingssemantichabit.Weintroducewordsemanticsimilarity-basedprivateinformationretrievalalgorithm.Boguskeywordselectioncriteriaarethekeypointofthealgorithm.Thealgorithmchooseswordsemanticsimilarityastheboguskeywordselectioncriteria,requiringsemanticsimilaritybetweenboguskeywordwithtargetkeywordmeetscertainconditions.ThetimecomplexityofthealgorithmisO(k),wherekistheamountofboguskeywords.Theexperimentsshowthatthequeryresultqualityofthismodelimprovesandtheinformationentropydecreases,IncomparisonwiththeGooPirmodel,butthedeclineofinformationentropyiswithinasmallrate.Keywords:privateinformationretrieval,wordsemanticsimilarity,boguskeyword,queryquality目录摘要I