If you aspire to return more detailed search results than: “See this bundle of documents” to your users, you are talking about piercing the document barrier.
It’s a benefit to be able to search thousands of articles and get the top ten (10) or twenty (20) for some search but assuming an average of twelve (12) pages per article, I’m still left with between one hundred and twenty (120) and two hundred and forty (240) pages of material to read. Beats the hell out of the original thousands or hundreds of thousands of pages, but not be enough.
What if I could search for the latest graph research and the search results opted out of the traditional re-explanation of graphs that wastes space at the first of nearly every graph article? After all, anyone intentionally seeking out a published graph article probably has a lock on that detail. And if they don’t, the paragraphs wasted on explanation aren’t going to save them.
DOM defines a platform-neutral model for events and node trees.
I expect to see graph-based implementations out in force. Given the recent “discovery” by some people that graphs are “universal.” Having a single node is enough to have a graph under most definitions.
For your amusement from Glossary of graph theory
An edgeless graph or empty graph or null graph is a graph with zero or more vertices, but no edges. The empty graph or null graph may also be the graph with no vertices and no edges. If it is a graph with no edges and any number n of vertices, it may be called the null graph on n vertices. (There is no consistency at all in the literature.)
I don’t find the lack of “consistency” in the literature surprising.