信息檢索算法在Web中的應(yīng)用與研究.pdf

1、湖南大學(xué)碩士學(xué)位論文信息檢索算法在Web中的應(yīng)用與研究姓名：岳文申請(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：陳治平20060325信息檢索算法在web中的應(yīng)用與研究AbstractWiththerapiddeVelopmentoftheInternet，WebhasbecomeanimportantwaytofindinformationSincetheonlineinformationkeepsincreasingrapidly，

2、howtoimprovetheef詫ctivenessandemciencyofinformationret“eValhasbecomethemostchaUengingproblemforsearchenginesInformationretrieValistheprocesstofindmorerelevantinformationforuser’srequirementinacollectionofdocumentsThispap

3、ermainlyconductsresearchonthealgorithmsofinformationretrievalbasedonsearchenginesFirstly，weintroducedthecurrentresearchprogressandreleVanttechnologyininformationretrieval，andcomparedtheperformanceofthreeclassicalmodelsfo

4、rinformationretrievalsystematicallyWealsoinvestigatedthedeVelopmentandsomeprimarymodelsofpersonalizedinformationretrieval，whichisbecomingmoreandmorepopularwiththerequirementsofpersonalizedandintelligentserVicesSecondly’w

5、iththeobservationthatveryshortqueriesininformationsearchingoftenTesultindepressedprecision，wepIoposedanovelinfbrmationretrievalalgorithmbasedonqueryexpansionandclassincationOurapproachattemptstocatchmorereleVantdocuments

6、byqueryexpansionandtextclassi6cationTheresultsofexperimentsshowthatthealgor“hmweproposedismorepreciseandemcientthanthetraditionalqueryexpansionmethodsFurthermore，weproposedapersonalizedsearchalgorithm，whichcombinesconten

7、tbasedfilteringandtextclassification，tosOlvetheproblemthattraditionalinformationretrievaltechnologiescan’tsatisfyanyqueryfromthedifferentbackgroundwiththedifferentintentionandatthedif托renttimeInourapproach，user’sinterest

8、modelisupdatedbymachine1earning，basedontheobserVationofuser’sbehaviorwhentheuserisbrowsingwebpagesWiththecontinuousupdate，user’sinterestmodelisbecomingmoreandmoreaccuratetouser’sinterestThus，theinformationinwhichuserisre

9、allyinterestedcanberetrievedeffectiVelybycontent—basedfilteringandtextclassificationaccordingtouser’sinterestmodelTheresultsofexperimentsshowthatthenewalgorithmismoreeffectivecomparedwiththetl’aditionalinfbrmationretriev

10、almethodsFinally，aprototypeofinformationretrievalwasdesignedandimplemented，basedontheabovealgorithmsandtechniquesKeywords：InfIormationRetrieVal；VectorSpaceModel；queryexpansion；textclassification；userprofile；contentfilte“