Figure 1
From: Dynamics of social network structure for Alzheimer and Lymphoma scientific communities
Unique structure dynamics for AD and Lymphoma social networks. (a) All PubMed publications of the last 16 ears, for each of the scientific field, were scanned. A list of either the last or the first author was prepared. Based on this list a connection matrix was built, with the number of joint publications for each researcher pair. Based on this connection matrix a neural network was built as explained in Horvath, Steve [4]. (b) For each network, the node degree distribution is plotted in a log scale. Note stable distribution for AD and an increase with time for Lymphoma (Lymph.). Kolmogorov-Smirnov (KS) test P value <0.05 for Lympohme epoch 1 and all other epochs and is not significant for any of the AD epoch comparisons. (c) For each network, the Weighted Cluster Coefficient (WCC) is plotted in time. KS test P value <0.05 between AD and Lymphoma WCC values. (d) AD social network is characterized by high density and cliquishness while the Lymphoma network is characterized by sparseness and low cliquishness. (e) Inside-country-WCC divided by the outside-country-WCC is plotted in time for different countries.