CORPUS. We have analysed [NUM] bibliographic records extracted from [ISI Web of Science/Scopus] and published from [FIRST YEAR] and [LAST YEAR]. Specifically, or corpus contained
ITERATION ON “YEAR”
NUM records published in 1st-YEAR-VALUE (ex. 123 records published in 2019)
NUM records published in 2nd-YEAR-VALUE (ex. 123 records published in 2019)
…
Our corpus consisted of
ITERATION ON “TYPE”
NUM 1st TYPE-VALUE (ex. 123 articles)
NUM 2nd TYPE-VALUE (ex. 234 conference proceedings)
…
BASE MAP. We extracted the [NUM] references present in this corpus and kept the [NUM] references cited by more than [NUM] records. We built the co-citation network of these references weighted by the frequency of their co-occurrence (aka bibliographic coupling). We removed all the reference-nodes with less than [NUM] connections. We spatialized the network with the ForceAtlas2 layout and fixed the position of the reference-nodes at equilibrium.
METADATA LAYER. From the same corpus we extracted and added to the network:
ITERATION ON “META VARIABLES”
NUM 1st META-VARIABLE-NAME occurring in more than N records (and with more than N connections)
NUM 2nd META-VARIABLE-NAME occurring in more than N records (and with more than N connections)
…
We connected these new nodes to the references co-appearing with them in the bibliographic records. We positioned new nodes using with the same layout algorithm while keeping fixed the position of the reference-nodes.
We sized the nodes the nodes according to the number of records in which they occurred and coloured them according to their type.
CORPUS. We have analysed [NUM] bibliographic records extracted from [ISI Web of Science/Scopus] and published from [FIRST YEAR] and [LAST YEAR]. Specifically, or corpus contained ITERATION ON “YEAR” NUM records published in 1st-YEAR-VALUE (ex. 123 records published in 2019) NUM records published in 2nd-YEAR-VALUE (ex. 123 records published in 2019) … Our corpus consisted of ITERATION ON “TYPE” NUM 1st TYPE-VALUE (ex. 123 articles) NUM 2nd TYPE-VALUE (ex. 234 conference proceedings) …
BASE MAP. We extracted the [NUM] references present in this corpus and kept the [NUM] references cited by more than [NUM] records. We built the co-citation network of these references weighted by the frequency of their co-occurrence (aka bibliographic coupling). We removed all the reference-nodes with less than [NUM] connections. We spatialized the network with the ForceAtlas2 layout and fixed the position of the reference-nodes at equilibrium.
METADATA LAYER. From the same corpus we extracted and added to the network: ITERATION ON “META VARIABLES” NUM 1st META-VARIABLE-NAME occurring in more than N records (and with more than N connections) NUM 2nd META-VARIABLE-NAME occurring in more than N records (and with more than N connections) … We connected these new nodes to the references co-appearing with them in the bibliographic records. We positioned new nodes using with the same layout algorithm while keeping fixed the position of the reference-nodes. We sized the nodes the nodes according to the number of records in which they occurred and coloured them according to their type.