I am not sure how the nodes size is calculated. I did a co-term network analysis using the default setting (scale with their weight and cooc-sum). In the documentation says:
Node size scale with their weight – The number of edges of a node determines its size
But, in my graph a node that has only one edge (travel community) is bigger than one that has 20 edges (selected node)
here a pic
Hope you can help me,
By default, node sizes are proportional to the number of cooccurrences they have with other nodes, which is not necessarily correlated with its number of connection in the final graph. By default, homogeneous network like co-terms are built using a distributional measure that links terms which share the same contexts. Therefore, a term with very high frequency but neutral meaning may show as a large node in the map with only few connections.
Note also that one can choose to represent the global frequency and not its total number of cooccurrences as node weight in the advanced node options of the mapping script. However the two measures are likely to be very correlated.
Thank you a lot, I did already the network using the frequency and the result is nearly the same. I am using this analysis for my PhD thesis, it is why I will really appreciate if you can tell me where I can find detail information about how the cooc sum is calculated.
thank you in advance.
It is just the total sum of the number of cooccurrences of a given term with any other term in the corpus!