運(yùn)用親和距離最優(yōu)排序聚類方法進(jìn)行期刊互引分析.pdf

1、中國(guó)科學(xué)技術(shù)信息研究所碩士學(xué)位論文運(yùn)用親和距離最優(yōu)排序聚類方法進(jìn)行期刊互引分析姓名：姜山申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：情報(bào)學(xué)指導(dǎo)教師：黃振中2001.9.1UseOptimalRelativeDifferenceArrangementinJournaltoJournalMutualCitationAnalysisABSTRACTAnoptimizedmethodwasusedtoanalyzejournaltojournalcrosscita

2、tionInajournaljoumalcitationmatrix，citingjournalsalerows，citedjoumalsalecolumns，andanelementisfrequencyofrowcitingcolumnDefineSk=∑∑cFIiJk=1～玎!扭l』=1whereiisarowjisacolumn，cijisallelementinthematrix，SkisthesumofallcO，IijIi

3、snamed“relativedifference”，kmeansthekⅡ1arrangementofrowsandcolumnsSupposewhenS。=vainS，七=l～ntthearrangementistheoptimalonetepresentthemutualityamongjoumalsThemethodislikeandmorecomplicatedthantheMinLA(MinimumLinealArrange

4、ment)problem，whichhadbeenprovedaNPcompleteproblemThemethodistemporarilynamedas‘‘OptimalRelativeDifferenceArrangement”UsuallyattempteachlocaloptimalalgorithmegSimulatedAnnealingAlgorithmtogaintheapproximateresultsoftheNP—

5、completeproblemsThispaperusedenumerationtogaintheoptimalarrangementMainworks：ProgrammedwitllenumerationalgorithmtoactualizeOptimalRelativeDifferenceArrangementUsedthemethodinjoumaltojournalmutualcitationanalysisCitationd

6、ataweregainedfromCSTPCD(ChineseScientificandTechnicalPapersandCitationsDatabase)publishedbyISTIC，andjournalswereselectedfromthedatableByanalyzingmanysamples，gotsomepositiveconclusionThemethodcanbetriedmoreinmumalcitation