Affiliationwww.cnnic.cn

InternationalConferenceonSocialScienceandManagement(ICSSM2014)432ResearchTrendOverviewforChinese-EnglishCrossLanguageInformationRetrievalJuanWang1,Peng-huiLyu2,*andXue-jingMa31LibraryofXiangnanCollege,Chenzhou423000,China;2SchoolofInformationManagement,WuhanUniversity.
Wuhan430072,China;3SchoolofInformationResourcesManagement,RenminUniversityofChina,Beijing,100872,China*email:lvph@whu.
edu.
cnKeywords:Chinese-English,CrossLanguageInformationRetrieval(CLIR).
Abstract.
ThepurposeofthisresearcharticleistoprovidesomereferencevaluesforfutureretrievalresearchonEnglishandChinesecross-languageinformationretrievalbasedonthedatafromWOSandCNKI.
BasedontheEnglishandChinesecross-languageinformationretrievalresults,acomparativestudyisintroducedandanumberofrelevantpapersinEnglishcross-languageretrievalfromdomesticandinternationalinstitutesweresignificantlyanalysed,thehigh-impactpaperswerestudiedusingitscontent.
ThisstudysummarizedthecurrentresearchfocusandresearcheffortsonthebasisoftheintroductiontothestatusinEnglishandChinesecross-languageinformationretrievalresearchathomeandabroad,finallyitcombedthemethodsofEnglishandChinesecross-languageinformationretrievalandthemainproblemstobesolvedandprovideareferenceforfurtherin-depthstudy.
IntroductionInrecentyears,thedemandsforpeople'sinformationsearchbecomeincreasinglystronger.
Theprogressanddevelopmentoftheinformationretrievaltechnologysuchastheemergenceofsearchenginesandnaturallanguagetextprocessinghavelargelyalleviatedtheaboveinformationretrievaldemands.
Afterthearrivaloftheeraof21stcentury'sglobalization,variousinternationalcommunicationisbecomingmorefrequent,thecross-bordertradeandeconomiccooperationisincreasing,somoreandmoreenterprises,organizationsandindividualsneedtousemulti-languageinformationsourcesformanypurposes,whilemanyexistingsearchenginesandothernetworkinformationretrievaltoolsaremostlydesignedforonesinglelanguage,whichcan'tmeettheincreasinglymulti-lingualinformationretrievaldemand.
Wheninformationinmultiplelanguageneeded,factorssuchasinformationretrievalskillsandlimitedlanguagesmasteringfromusersalsorestrictthesatisfactionoftheirinformationneedsseverely,causingthempayinglotsofcostandprice.
Therefore,theproblemofinformationretrievalamongdifferentlanguageshasbecomeoneofthemajorobstaclesofglobalinternetusersnow,obtainingfurtherinformationtheyneed,thissituationisparticularlyprominentforinternetusersfromChinawiththetoppedworld'spopulation.
Accordingtothelatestdata,thefirstteninternetlanguageusers'accountsfor84.
3%ofthetotalnumber,ofwhichEnglishusersaccountsfor28.
7%ofthetotalnumber,theChineseuserscomesinthesecondplace,accountingfor21.
7%ofthetotalones,soEnglishandChinesehavebecomethetwomajorlanguageswiththelargestusers[1].
Bytheearly2012,thetotalnumberofinternetusersfromChinahasexceeded500million,reaching513million[2].
WithChina'snationalcomprehensivestrengthimproving,itsinternationalinfluenceisincreasingdaybyday,thoseexchangeandcooperationbetweenChinaandthewestinculture,academic,scientificresearchandmanyotherfieldsareincreasinglyclose,resultinginChineseandthewestusersboundtocontactheavilyanduseinformationaswellasknowledgeindifferentlanguages.
Therefore,theworkontheChineseandEnglishcross-languageinformationretrievalisverynecessaryandhasgreatpracticalsignificance.
InternationalConferenceonSocialScienceandManagement(ICSSM2014)433Thegeneralprocessofinformationretrievalis:automaticindexingsystem,artificialindexingindexes,andsomecharacteristicsofthedocumentliteratureareregardedasliteratureidentification;whenusersneedliteraturesonatopic,theyexpresstheirinformationneedsbyconstructingretrievalexpression,andsubmitthemtotheretrievalsystem.
Retrievalsystempresentssearchresultstotheusersthroughcertainenumerationalgorithm.
Incross-languageinformationretrieval,retrievalexpressionandliteratureidentificationaretwodifferentlanguages,soitneedstoconverttothesamelanguageandthenusematchingprocessing;andconvertingandmatchingbetweendifferentlanguagesarethemaindifficultiesincross-languageinformationretrieval.
Fig.
1:NumberofCLIRathomeChinaandabroadTheso-calledChinese-Englishcross-languageinformationretrievalreferstotheactivitiesthattheusersconstructretrievalexpressioninChineseorEnglishtoretrievetheChineseorEnglishliteratureinretrievalsystem.
Chinese-Englishcross-languageinformationretrievalresearchaimstoprovideanoptimalsolutionplatformforinformationretrievalproblembetweenthetwodifferentlanguagesofChineseandEnglish.
Inthiswayusersconstructtheretrievalexpressionwiththelanguagewhichtheyaremorefamiliarwith.
Thiscaneliminatetheinformationretrievaldifficultiescausedbylanguagebarrierstoacertainextent,soastosolvetheproblemofChineseandEnglishinformationretrievalandeffectivelyresolvetheproblemofcross-languageretrievalbetweenthetwolanguagesusedbythelargestnumberofworldwidepeople.
WhatistheresearchsituationonthisproblemathomeandabroadBysearchingforeignliteraturedatabasesrepresentedbyWebofScience(WOS)fully[3]anddomesticliteraturedatabasesrepresentedbyNationalKnowledgeInfrastructure(CNKI),thisarticlefindsthatinternationalresearchonthistechnologystartedearlier,andhashadmanyresearchresultsbefore2005asFig.
1shows.
Despitestartinglater,thegrowthofthedomesticresearchissoquick,thepaperamountresultsafter2007ismorethanthatofforeignandincreaseaheadforalongtime,itcanbelearnedthatontheonehand,domesticdemandforChinese-Englishcross-languageretrievalismoreurgent,ontheotherhand,domesticinvestmentismoregreatthanthatofforeignintheprogressonthetechnologyandthusgainssignificant.
Ofcourse,theadvantageonamountisonlyoneaspectreflectingproblemssolving.
Thisarticlemakescomparisononquality,textandtechnologyinterpretationthroughmostcitedpaperbelow.
ThelowersolutionbasedoninterpolationInternationalResearchTable1listsoftheforeignhigh-impactpapersachievementsTimeCitedTitleAuthorAffiliationResourcesPublicationYear24TerrierinformationretrievalUnivGlasgow,AdvancesIn2005InternationalConferenceonSocialScienceandManagement(ICSSM2014)434platformScotlandInformationRetrieval20AnchortextminingfortranslationofWebqueries:AtransitivetranslationapproachAcadSinica,InstInformatSci,TaiwanACMTransactionsOnInformationSystems200420Combiningmultiplestrategiesforeffectivemonolingualandcross-languageretrievalUnivNeuchatel,InstInterfacInformat,SwitzerlandInformationRetrieval200420MorphologicaltypologyoflanguagesforIRTampereUniv,DeptInformatStudies,FinlandJournalOfDocumentation2001Globalcross-languageinformationretrievalresearchhassomeregionalcharacteristics,theU.
S.
andEuropeanresearchersaimatresearchandpracticeinEnglish,German,French,SpanishandotherEuropeanlanguagesmainly.
ComparedtoChinese-Englishcross-languageinformationretrievalsystem,CLIRsystembetweenthemisrelativelyeasytoimplement.
ButChinese-Englishcross-languageinformationretrievalresearchresultofforeignresearchersgetsgreatimpactduetoeconomicandtradeactivities,theinternationalfinancialcrisisinrecentyearshasledtothedeclineofattentionoftheresearchindirectly,however,someinternationallyrenownedconferenceonCrossLanguageInformationRetrieval(CLIRConference)[4],thesubjectsortopicsofthemeetinginrecentyearsarealsomoreorlessrelatedtotheChinese-Englishcross-languageinformationretrieval,whicharereflectedclearlyinthepublicationformofresearchresultsabroad.
Fromtheinformationlistabove,thefollowinginformationofmajorcross-languageretrievalinternationalconferencescanbesummarized:AdvancesinCross-LanguageInformationRetrievalisthemostinfluentialinternationalanthologyonCLIR.
MostinfluentialworksontheCLIRinterrierinformationretrievalplatformispublishedthroughthejournals.
InternationalCLIRresearchcentersaremainlyconcentratedinsomecollegesanduniversitiesintheUnitedStatesanditisthemthatleadtheadvancementofCLIRcutting-edgetechnology.
Inaddition,thisarticlehasalsonotedsomeinfluentialCLIRinternationalconferenceheldconsecutively,suchastheinternationalconferenceslistedbelowonCLIR:(1)TextRetrievalConference(TREC)TextRetrievalConference(TREC)isoneoftheworld'sauthoritativeannualevaluationconferencesinthefieldofinformationretrievalresearch.
In1992,theU.
S.
DefenseAdvancedResearchProjectsAgencyandtheU.
S.
NationalBureauofStandardsandTechnologyhostedTextRetrievalConferencejointlyandCLIRisoneoftheimportanttopicsdiscussedinTREC.
ThepreviousTRECfocusedonCLIRmainlybetweenwesternlanguages,butinrecentyearsithasaddedChinese,ArabicandEnglishcross-languageretrievalevaluationproject,therelevantinformationcanbefoundathttp://trec.
nist.
gov/.
(2)Cross-LanguageEvaluationForum(CLEF)CrossLanguageEvaluationForum(CLEF)isoneofthecontentsofadigitallibraryresearchfundedbytheEuropeanCommission,focusedonresearchanddiscussionofmonolingualretrievalmainly,cross-languageandmulti-languageretrieval,restrictedareasretrievalandinteractiveretrievalbetweenEuropeanlanguages,therelevantinformationcanbefoundinitsofficialwebsitehttp://clef.
isti.
cnr.
it/.
(3)JapanNationalScienceInformationSystemsCenterInformationRetrievalSystemTestSetMeeting(NTCIR)NII-NACSISTextCollectionforIRSystems(NTCIR)isaninformationretrievaltestsetevaluationconferenceorganizedbyJapan'sNationalInstituteofInformatics,themeetingmainlyconstructsAsianlanguages(Chinese,Japanese,Korean)relativetestdocumentsetsandmakestheevaluationonrelatedretrievalsystem,testitemsincludesingle-languageinformationretrieval,cross-languageinformationretrieval,question-answeringsystem,textsummarizationandwebsearch,specificinformationcanbefoundathttp://research.
nii.
ac.
jp/ntcir/index-en.
html.
InternationalConferenceonSocialScienceandManagement(ICSSM2014)435ChinaDomesticResearchTheChinadomestichigh-impactresultsaremainlypublishedintheacademicjournalsofLibraryandInformationScience(LIS)field,whichisdifferentfromtheinternationalininfluentialconferencesinindustry.
MostoftheresultsarealsodistributedinacademicjournalsinthefieldofLIScategory,whichshowsthattheattentiononChinese-Englishcross-languageretrievalaremainlyfromthefieldsofLIS,aswellasfieldofinformationmanagement.
Onethirdofthesehigh-impactpaperscomefromthePekingUniversity,itcanbelearnedthatChina'scross-languageretrievalresearchcentersaremainlyinthemainlandChina,sodomesticuniversitiesalsohavealotofhigh-qualityresearchoutputs.
Table2listsofthedomestichigh-impactpapersachievementsTimeCitedTitleAuthorAffiliationResourcePublicationYear35English-ChineseCross-languageInformationRetrievalUsingLuceneSystemDepartmentofComputerScience&Engineering,FudanUniversityComputerEngineering200516BilingualDictionaryExtractionforSpecialDomainBasedonWebDataOpenSystemandChineseInformationProcessingCenterInstituteofSoftwareChineseAcademyofSciencesJournalofChineseInformationProcessing200612DevelopmentandOutlookofCross-languageInformationRetrievalDepartmentofInformationManagementofPekingUniversityResearchesinLibraryScience200612AnEffectiveMethodtoExtractTranslationPairsfromWebCorporaSchoolofSoftwareandMicroelectronicsofPekingUniversityJournalofChineseInformationProcessing2008Aftercarefullyanalyzingandinterpretingtheabovepapertextslistedandnotlisted,thispapersummarizesanddescribestheirresearchprogressandthemainpointasfollows:XiaoJianfromChinaElectronicInformationIndustryDevelopmentResearchInstitutedeemedthat:multi-wordexpression(MWE)canimprovethequalityofcurrentmachinetranslationsystem,whichcanimprovetheeffectofcross-languageinformationretrieval[5].
Afterexperimentalsimulation,theyprovedthatusingthemethodbasedonsemanticandstatisticalextractandmulti-word,theexpressislargelyfeasible,andEnglish-ChineseMWEtranslationcanbeextractedfromthetriplescomparablecorpus,soastoprovidevaluableresourcestocross-languageinformationretrieval.
WuDanfromWuhanUniversityhasmorecomprehensivestudyoncross-languageinformationretrieval[6],basedonyearsofresearchincross-languageretrieval,shepublishedhismonograph"cross-languageinformationretrieval-translationoptimizationtheoryandtechniques"in2011.
ForquerytranslationambiguityproblemofCLIR,sheusedaninteractivedevelopmentsystemanddesignmethod,basedonrelevantfeedbackofcross-languageinformationretrievaltechnology,presentsanEnglish-Chineseinteractivecross-languageinformationretrievalsystem,herexperimentalresultsshowthatthesystemimprovestheretrievaleffectsignificantly.
XiaLixin,fromCentralChinaNormalUniversity,inviewofubiquitousshortagesuchaspoortranslationaccuracy,lowefficiencyandhighcostinexistingcross-languageretrievalmodel,putforwardanEnglishandChinesecross-languageretrievalmodelbasedontopicmap[7].
Translationoftopicmapisfulfilledinthecontextguidanceofrichrelationshipandscopeofthetopicmap,whichovercomesthetranslationambiguityproblemofquestion-translationstrategycausedbylackingofcontext.
Intheprocessofretrieval,throughthetopicmapitmakessemanticprocessingtoretrieveobjects,analyzesthesemanticrelevanceofretrievedobjectsandqueryrequest,andfinallymakesaInternationalConferenceonSocialScienceandManagement(ICSSM2014)436judgmentwhetherasaresultofthereturn.
Outstandingfeaturesofthismodelarethatitcanimprovetheaccuracyofthetranslationandtheeffectivenessofretrieval.
ZhangYuejiefromFudanUniversityusequerytranslationbasedonmachinetranslationaslanguageboundariesacrossthesourcelanguageandthetargetlanguage,anduseknowledgesourceswhichEnglish-Chinesebilingualdictionariesasthemainbodytocompletethequerytranslationprocess,establishedEnglish-ChineseCross-LanguageinformationretrievalsystemintheTREC-9[8].
InordertofindamoreeffectivemethodofquerytranslationbetweenEnglishandChinese,aswellasimprovetheperformanceofChineseretrievalsystem,theyestablishedmonolingualChinesesearchenginebasedontoolofLucenein2005,andputforwardtheirownalgorithmsinEnglish-Chinesequerytranslationandevaluationsystems[9].
In2009,basedontheevaluationoftextretrievalmeetingonEnglish-Chinesecross-languageinformationretrieval,theytookEnglish-Chinesebi-directionalquerytranslationandEnglishtoChinesetranslationasdominantstrategyandtranslationobjectrespectively,usingEnglish-Chineseelectronicdictionaryasknowledgesourceofacquiringtranslationknowledge,combiningwithbuiltEnglish-Chinesemonolingualinformationretrievalsystems,achievedcompleteEnglish-Chinesebi-directionalcross-languageinformationretrievalprocess[10].
Ofcourse,inadditiontoacademicjournalpapers,thispapernotedthatZhangDongweifromHeilongjiangUniversityhassystematicallydiscussedtheprobleminhis2006dissertation,andbuiltamodelinquestionprocessingstage,basedonwordssegmentationmethodandcorpusmethod,cuttingquestiontypeintowordswithpracticalsignificance.
Intheliteratureprocessingstage,thosemethodswereemployed:usingnaturallanguagemethodsautomaticindexingdocumentcontents,usingmethodsofwordfrequencystatisticstoextractwords,afterprocessingoftranslationbeingaddedintothedictionaryofsystem;usingdifferentlinguisticresourcesimprovingdictionaries,andusinghomologymatchingstrategyfortheprocessofpropernouns.
Theinnovationofthemodelistheestablishmentofknowledgedictionarybasedonfilecontent,reducedtheuncertaintyofthetranslationprocess,andimprovedthequalityoftranslation.
ConclusionsFig.
2:classificationofcross-languageinformationretrievalmethodsForthefutureresearchdirectionoftheChinese-Englishcross-languageretrievaltechnology,thispaperinsistthatthemostpromisingthingistheuseofontologytoestablishthecorrespondingcorpuswhichisclosertothesemanticnatureoftheinformationretrievaltechnology.
Becausethebilingualontologycanportraycorrespondingdomainknowledgeofdifferentlanguages,soitcansolvesemanticlossanddistortionproblemsemergingintheconversionfromonelanguagetoanother.
AlthoughtherearemanymorematureEnglishontology,whiledomesticresearchonontologyisstillinitsinfancystage.
IfturningmorematureEnglishontologyintoEnglish-Chineseontologyneed,forInternationalConferenceonSocialScienceandManagement(ICSSM2014)437Chinese-Englishcross-languageinformationretrievalresearch,itwillhavemeaningsasmilestone.
ForEnglishontologybeingturnedintoEnglish-Chineseontology,theproblemfacedistheEnglish-Chinesetranslation.
HuXinyanfromOceanUniversityofChinaproposedthemanagementoftheChinese-EnglishontologybasedonChinese-EnglishBilingualCorpus(CEBC)[11],andhasmadeagoodexplorework.
Afterreviewworkthispaperarguesthatmethodsandtechniquescommonlyusedincross-languageinformationretrievalcanbeclassifiedasshowninFigure2.
Duetodocumenttranslationreferstoalltheliteratureorliteratureidentificationofretrievalsystem,theyaretranslatedintothesamesourcelanguageformofretrievequestionformula,thusbecausethereisamorecompleteliteraturecontext,itcouldachieveacertainsemanticdisambiguationinthetranslationprocess.
Querytranslationreferstotranslatingretrievequestionformulaenteredintothelanguagebytheusersupportedbythesystem,whiletheintermediatelanguagetranslationreferstobothretrievequestionformulaandliteraturearetranslatedintounifiedintermediatelanguage.
Finally,methodsofachievecross-languageinformationretrievalwithouttranslationaremainlyhomologousmatching,latentsemanticindexingaswellasgeneralizedvectorspacemodelsandsoon.
ItcanbeseenfromFig2thetranslationbetweendifferentlanguagesisthekeyproblemofcross-languageinformationretrievaltechnology,theleveloftranslationqualityaffectsretrieveeffectsofcross-languageinformationretrievalsystemdirectly.
InChinese-Englishcross-languageinformationretrieval,themainproblemsoftranslationarethefollowing:(1)Translationambiguityproblems.
AChinesecharactercanbetranslatedintoanumberofEnglishwords,whileanEnglishwordhavemorethanoneChinesemeaningitself,sotherewillbetranslationambiguityprobleminthetranslationprocess.
AccordingtoYuHuijia'sstudybasedonalarge-scaleuserssearchinglogs[11],theaveragequerylengthspentbythedomesticsearchengineusersis1.
85words,93.
15%ofthequeriesarenotmorethanthreewords.
Therefore,fortheuserstheygiveashortretrievalquestionformula,whichvirtuallyincreasesthedifficultyofeliminateambiguityandquerytranslation,certainquestionsformulamustbeextendedinordertogetabettersearchresults.
(2)Unknownwordproblems.
Indictionary-basedChinese-Englishcross-languageinformationretrievalmethods,itoftenencounteredproblemsofthequestionwordcontainingnewwords,phrasesofspecificacademicfieldsorpropernounwhichbeyondthescopeofthelexicon,recognitionproblemofnewwords,shortwords,foreignwords,propernounscausesnodifficultytoquerytranslation.
(3)SomecharacteristicsofChineseownalsohavecausednosmallobstacletoChinese-Englishcross-languageinformationretrievalUnlikeEnglishwords,thereisnoseparatorinChinese,soitshouldmakesegmentationforthecharactersbeforeinformationretrieval.
SegmentationalgorithmofChinesecharacterhasadirectimpactonthetranslationeffectsofretrievalquestionformula,althoughalotofresearcheshavebeendoneontheChinesecharactersegmentation,therearestillmanyproblemstoberesolved.
Chinesesyntacticandsemanticanalysisareextremelycomplex,additionalinformationofChinesegrammarisrelativelyfew,thereisnochangeofpartsofspeechandsingularandpluralvariations,intheprocessoflanguageprocessing,allthesereasonshavebroughtmanydifficultiestoproceduresautomaticidentificationandtranslationsbasedonChinese-Englishdictionary.
However,duetotoomanyusersofChineseandEnglishexited,English-Chinesecross-languageinformationretrievalhasspecialimportantsignificance.
ThispaperhasdoneasimplecombingandthinkingforcorrelationalresearchofChinese-Englishcross-languageinformationretrieval,themethodofliteraturesearchandreviewbasedonthenumberofhighlycitedpapersprovidedfromthispaperalsoprovidedapathofreferenceforresearchcommentaryandsummaryofotherrelatedfields,itisaneffectiveexampleofinformationretrievaltechnologyappliedintheacademicresearch.
Ofcourse,forthesometechnicalproblemsandresearchquestionsproposedinthispaper,itwillcontinuetoexploreinourfutureresearch.
Meanwhile,withmoreandmoreattentionandresearchofInternationalConferenceonSocialScienceandManagement(ICSSM2014)438scholars,itcanbebelievedinthefieldofChinese-Englishcross-languageinformationretrievaltherewillhaveabiggerbreakthroughandlong-termprogressinnearfuture.
AcknowledgementTheresearchworkwassupportedbyNationalScienceFoundationofChinaunderGrantNo.
71303090.
TheauthorsaregratefultoXiangLiu,RenDingandJinZhangfortheirhelpfuldiscussionsandsuggestions.
Andourspecialthanksalsodedicatetothoseanonymousreviewersfortheirvaluablecommentsforthisresearchwork.
References[1]http://internetworldstats.
com/stats7.
htm.
[EB].
[2013-12-1].
[2]ChinaInternetNetworkInformationCenter.
The29thChinaInternetnetworkdevelopmentstatestatisticsreport.
[EB].
[2013-12-01].
http://www.
cnnic.
cn/dtygg/dtgg/201201/W020120116337628870651.
pdf.
[3]Ma,Feicheng;Lv,Penghui;Ye,Mao.
StudyonGlobalScienceandSocialScienceEntropyResearchTrend.
IEEE5thInternationalConferenceonAdvancedComputationalIntelligence.
pp.
238-242,2012.
[4]ZhangDongwei.
ModelresearchonChinese-Englishcross-languageinformationretrieval,HeilongjiangUniversityPress,2006.
[5]XiaoJian,Xujian,XuXiaolan,YuanQi.
AutomaticextractionandalignmentofmultiwordexpressionsfromEnglish-Chinesecomparablecorpus,(31),pp.
130-134,187,2010.
[6]WuDan.
DesignandimplementationofanEnglish-Chineseinteractivecross-Languageinformationretrievalsystem.
NewTechnologyofLibraryandInformationService,(2),pp.
89-95,2009.
[7]XiaLixin,WangZhongyi.
TheconstructionofChinese-Englishcross-languageinformationretrievalmodelbasedonthetechnologyoftopicmap.
LibraryandInformationService,(11),pp.
70-74,2008.
[8]ZhangYuejie,GuoYikun,Lianli,WuLide.
Implementingcross-LanguageinformationretrievalbasedonEnglish-Chinesemachinetranslation.
Mini-microSystems,(7),pp.
1135-1140,2004.
[9]ChengShijie,ZhangYuejie.
English-Chinesecross-languageinformationretrievalusingLucenesystem.
ComputerEngineering,(13),pp.
62-64,2005[10]Yanghui,ZhangYuejie,ZhangTao.
Dictionary-basedmethodforEnglish-Chinesebidirectionalcross-Languageinformationretrieval.
ComputerEngineering,(16),pp.
273-274,277,2009.
[11]HuangXinyan,YaoWenlin,XuJianliang.
EstablishmentandmanagementoftheChinese-EnglishOntologybasedonthebilingualcorpus.
ChineseJournalofScientificInstrument,(S2),pp.
529-532,540,2005.
[12]YuHuijia,LiuYiqun,ZhangMin,RuLiyun,MaShaoping.
Researchinsearchengineuserbehaviorbasedonloganalysis.
JournalofChineseInformationProcessing,(1),pp.
109-114,2007.