📖Notes on: “Graphs, Metagraphs, RAM, CPU” by Linas Vepštas
- authors
- vepstas2023-graph-metag-ram-cpu
- year
- 2023
- url
- https://github.com/opencog/atomspace/blob/master/opencog/sheaf/docs/ram-cpu.pdf
- reading status
done
Hypergraphs: edges/links are ordered list of vertices/nodes (0 or more arity)
(This can be used to “type” edges/links. As otherwise, edges have no distinct identity beside a list of targets.)
Metagraphs: edges/links are allowed to reference other edges/links
Graphs can be represented as metagraphs, and metagraphs can be represented as graphs. However, storing graphs as metagraphs incurs a small overhead, while storing metagraphs as graphs has a high overhead.
NoSQL is a dual of SQL (coSQL). References: Erik Meijer and Gavin Bierman, “A co-Relational Model of Data for Large Shared Data Banks”, ACM Databases, 9, 2011, URL https://queue.acm.org/detail.cfm?id=1961297.
Most query languages are implicitly designed around tables (drawing inspiration from SQL) and are limiting their usefulness as a graph query language.
p.39 Sapir–Whorf Hypothesis: language influences speakers’ worldview and cognition. In the same way, the query language limits how you approach the problem.