Topics in Research on International Relations as Clusters of Citation Links

"journals"
##########
* list of full and abbreviated names of journals used for downloads from Web of Science
** journals.csv


"data cleaning"
###############
* list of identified reference strings
** ref.sou.csv
* code for reference identification (by Felix Mattes)
** identify.references-revised.R


"network"
#########
* bipartite network of 
** top-300 highly cited sources
** 6,494 papers in IR journals and books 2011-2015 citing at least two top-300 sources

* data.frame sel.sou of selected top-300 sources (ordered by sou.id)
** columns:        
[1] sou.id		= source id
[2] fam.aut.1   = family name of first author
[3] n.cit       = number of citations 2006-2015 in expanded data set
[4] year        = year of first publication
[5] ref         = typical reference
[6] co.aut      = co-authors
[7] title       = title
[8] journal     = journal (of journal papers)
[9] journal.abbr= abbreviation of journal
[10] IR         = classified as IR 
[11] v          = volume
[12] p          = first page 
[13] sn.aut.1   = surname of first author 
[14] c.rank     = citation rank of n.cit 
[15] r.n.cit    = reduced number of citations (reduced to IR journals and books) 
[16] r.c.rank   = citation rank of r.n.cit (ties broken by year)
** sel.sou-v12.RObj

* number of
** nodes: n = 6,494 + 300
** links: m = 30,835

* link list lili of citation links 2011-2015 from 6,494 papers to top-300 sources (and reverse)
** ids of cited sources: 1:300 in order of sel.sou
** ids of citing papers: 300 + 1:6494
** columns:        
[1] n.1	= node 1 (id of citing paper in first m rows and of cited top-300 source in last m rows)
[2] n.2 = node 2 (vice versa)
** lili.2011-2015-v2.300.RObj


"Ward clusters of views"
########################
* made by Michael Heinz
* used as seeds for link clustering
** cv.RObj


"PsiMinL"
#########
* R-package for memetic link clustering programmed by Andreas Prescher
** will be delivered by request
* parameters of experiments in IR studies in R-scripts:
** adapt-Ward-seed-v5.R
** graph-v2.R
** run-phases-v4.R
** parameters-v2.R

"link clusters"
###############

* logical matrix v.Ls of 19 valid link clusters and m links
** names of 19 columns	= names of clusters (cf. paper)
** names of m rows		= number of links (parallel to lili[1:m,])
** v.Ls-v6.RObj

* numerical matrix ccs of cluster citations of top-300 sources
** contains internal degree of each top-300 source in each valid cluster and in the whole network (all) 
** names of 21 columns	= names of clusters (cf. paper) + all  + IR (classication as in sel.sou)
** names of 300 rows	= family name of first author and year of first publication (parallel to sel.sou)
** ccs-v7.RObj

* R-code for getting 
** full-member sources of a link cluster lc: 
which(ccs[,lc]==ccs[,"all"])
** sources of a link cluster lc with membership grade larger than mg < 1: 
which(ccs[,lc] > mg * ccs[,"all"])
** citation rank-list of sources with regard to internal degree in cluster lc 
rev(sort(ccs[ccs[,lc] > 0, lc]))