分類概念語義特征構(gòu)建技術(shù)的研究.pdf

1、東北大學(xué)碩士學(xué)位論文分類概念語義特征構(gòu)建技術(shù)的研究姓名：趙研申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：計(jì)算機(jī)應(yīng)用技術(shù)指導(dǎo)教師：高克寧20070101東北大學(xué)碩士學(xué)位論文AbstractResearchTechnologyforConstructingSemanticFeatureofClassificationConceptAbstractAsInternetdevelopingrapidlyandbeingmoreandmorepopularWebha

2、salreadybecomeahugeglobalinformationwarehousewithvarietiesoflesomcA葛availabletoshareandit’salsoaneffectiveandexcellentme鋤toobtaiuinformationforpeopleForWebl酷oureesbeingtremendousinamountandcompleteincategorynearlyanyinfo

3、rmationcanbefoundonitNeverthelessconfronted州ththehugeseathatisfullofinformationwhatisreallyusefulforagivenu孵occupiesalittlepartofitIt’simportanttoprovideaneffectivetoolassearchingengineorinformationintegrationsystem，help

4、ingpeopletofindthedesiredinformationthatthey眥interestedinproperlyandqmeHyIntheinkmenseinformationspaceWebinformationi8usuallyorganizedinformofWebsiteswhichestablishtlleirownclassificationcategoriesforpagesorganizationand

5、informationpublicationachievingthestructuralclassificationsystemwithlevelsHoweverfordifferentW曲6it韶theclassifiedcriterionisnotunifiedandtheydon’thavethestandardfornamingcategoryitemswhichleadstothedifferencesofsemanticsW

6、iththedifferences，multi剮MⅡ∞Webinformationcouldn’tsatisfythepurposeofcompatibilitymergereventheeffectiveintegrationAwayforunderstandinganddescribmgsemanticsofWebinformationisneededurgentlyTosettlethesemanticdi伍釉c懿ofcatego

7、ryitemsofWebsitesthecontributioninthisthesisisanalyzingclassificationmlesfromtwopointsoflevelstrucUueandinformationsemanticsofWebinformation，and#vmgouttheclassificationsystemforeachWebsiteForasingleclassificationsystem，t

8、heatomnodesofitareclassificationconcepts，whosesemanticfeaturesandlevelstructuresreflectthepot訓(xùn)alsemanticsandarrangementofWebcategoryitems，resolvingtheproblemofunifonnunderstandingandstandarddescriptionofdifferentWebcateg

9、oryinformationThemainworkforthethesisisaboutthepurposeofdescribmgthesemanticfeatureofWebcategorywhichconsistsofthreeaspects，thatareWebdataextractingandpretreatmentextractingfeatureinformationfromclassifiedWebtexts，andrep

10、resentingthesemanticfeatureofclassificationconceptsinformofvectorsThemainachievementsmdownloadingthetextsfromtheWebpages，analyzmgHTMLtags，extractingfeatureitemsthatcouldrepresentthepotentialcontentandtopicofthecategoryfr