word-streams for representing context in word maps

word-streams for representing context in word maps

ID:34381214

大小:86.15 KB

页数:6页

时间:2019-03-05

word-streams for representing context in word maps_第1页
word-streams for representing context in word maps_第2页
word-streams for representing context in word maps_第3页
word-streams for representing context in word maps_第4页
word-streams for representing context in word maps_第5页
资源描述:

《word-streams for representing context in word maps》由会员上传分享,免费在线阅读,更多相关内容在教育资源-天天文库

1、Word-StreamsforRepresentingContextinWordMapsArnulfoP.Azcarraga,AlfredKennethS.Gopez,andTeddyYapJr.ProgramforResearchintoIntelligentSystemsSchoolofComputing,NationalUniversityofSingapore10KentRidgeCrescent,Singapore119260email:{arnulfo+gopez+tyap}@comp.nus.edu.sgAbstractreceivedenormousatte

2、ntionasadevelopmentalareaofThemostprominentuseofSelf-OrganizingMapsresearch.(SOMs)intextarchivingandretrievalistheWEBSOM.TechnqiuesvarywidelyinthespecificprocessesInWEBSOM,amapisfirstusedtoreducetheinvolvedintextarchivingandretrieval,andtherehasdimensionalityofthehugetermfrequencytablebybe

3、enconsiderableresearchprogressinthevarioustrainingaso-calledword-categorymap.Thisword-approachestofeatureextractionanddocumentcategorymapisthenusedtoconverttheindividualclassification.Thesetechniquesspanthefieldsofdocumentsintotheirrespectivedocumentsignatures(i.e.ArtificialIntelligence,St

4、atisticalPatternRecognition,histogramofwords)whichformthebasisfortrainingaandArtificialNeuralNetworks.Yang[1]recentlydocumentmap.Thisdocumentmapisthefinaltextcomparedvariousgeneralmethodsfortextclassification.archive.WEBSOMhasbeenshowntobeapowerfulandBenchmarkcasesarenowavailabletocomparen

5、ewversatiletextarchivingsystem.However,itspendsmethodswiththeolderones.(wastes)enormouscomputerresourcesintheTheconditionsfortextclassificationhavedrasticallycomputationoftheleftandrightcontextofeachandchangedovertheyearsandthisexplainswhynewereverywordthatappearsinanyofthedocumentsintheme

6、thodsarecontinuallybeingconsidereddespitethetextcorpus.Thispaperpresentsanalternativeschemerelativeefficiencyoftheoldermethodsandthefactthatforincorporatingcontextintheencodingofthewordsinthesearealreadywellunderstood.BecauseoftheWeb,suchawaythatthecomputationoftheprobabilisticplusthefactt

7、hatpersonalcomputersandotherwidelycentroid,whichisinherentintheSOMtrainingavailablecomputingdeviceshavebecomesopowerful,algorithm,istakenfulladvantageof.Severaldigitallibrarieshaveventuredintoapplicationsthatwereexperimentsareconductedtocomparethisnewschemenot

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。