So, having had a bit of a play, a look at what people are doing, and just mucking around with data, I ended up essentially making some up just to see what I was trying to (in principle) do. So this [page]1 explaining how the data tables are laid out in gephi was useful. So as I understand, the keywords go in as ‘nodes’, and the edges are any time a keyword is connected to another – so that could be on a single article, or it could be across articles; the ‘edges’ could be labelled by article title (although I seem not to be managing that in a way that looks acceptable at the moment). So in this mock thing I: Had 5 articles, with sets of tags 1 tag applied to 4 of the articles – so as well as relationships (edges) between the tags (nodes) for those articles, edges also link them to others via that node. The 5th article doesn’t have an edge to that node (i.e., wasn’t tagged with it), but it does have a relationship with another tag used in another article. So I played around with the different types of layout…I also did some things that I couldn’t work out how to undo (hence, it’s all red) this probably looks most sensible(ish): []2 Now, obviously, to anyone who knows what they’re doing this is pretty pathetic – but as previously mentioned, I’m fundamentally a moron so small steps = at least some happiness.