I like the idea of interactive filters to allow the person to explore it visually. I'm hoping to have something people can play with in the next few weeks. I hadn't really expected the site to take off the way it has and so it's not really feature complete yet :)
But what you're discussing is a frequent problem with graph visualization - it's very easy to end up with 'hairball charts' that may be meaningful to the person who generated them but only because of familiarity with all the steps it took to get there, and the more inclusive the graph, the more time eaten by clustering algorithms, the more clusters produced, and the more of a cluster...well you get the idea.
As you're at an early stage of development, perhaps this technique, which trades away completeness for clarity and is relatively novel, might let you leapfrog some of those problems: https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=8896846
It is a nice user interface and the reference material is useful and well presented. But when you get down to it, it's linear lists about the characteristics of a semantic graph. As I've said many times, this is like describing a tree with a tabular catalog of its leaves. Graph navigation needs a graphical representation, because things like branchiness (node out-degree) and other factors are more easily shown than described.
The basic problem with graph representation/ navigation/ traversal is that there are many valid ways of looking at the graph and it's hard to render them all. Maybe try using a gutter to allow users to temporarily pin certain graph characteristics and render accordingly. In this context, sometimes I might be interested in the latest research that cites a paper, other times I might be looking to see who picked it up first, or to apply some sort of windowing function to a large spectrum of citations.
But I want to see the graph, even if I am looking for a particular leaf.