Categorical Informatics- A Functorial Approach to Data Integration
| Author | David I. Spivak |
|---|---|
| Published | 2015 |
| Format | Report |
| Original | Read the original |
TL;DR — David Spivak’s 2015 NSF grant proposal arguing that category theory should be the mathematical foundation for integrating data across databases.
The proposal, dated November 1, 2015, starts from a practical problem: data is produced under many different models, and combining it is costly. It cites an estimate that data integration takes about 40% of enterprise IT budgets and argues that existing tools are ad hoc because the field has no principled mathematical foundation. Category theory, “the mathematics of structure,” is offered as that foundation, building on five years of work at MIT and Harvard in which database schemas are treated as categories and data migration is expressed through functors. That work was implemented in software, including the FQL (Functorial Query Language) tool.
The proposed work has three topics: constraints, so that schemas can enforce rules and data can be repaired with clearer semantics than existing “chase” procedures; view updates, meaning how changes made to a derived view are pushed back to the underlying data; and nesting and aggregation, turning flat data into grouped data. It also lists industrial and government partners (NASA, NIST, Amgen, Janssen, Honeywell) and Spivak’s book Category Theory for the Sciences. The document follows the NSF format: project summary, project description, broader impacts, results from prior NSF support, references, biographical sketch and a postdoctoral mentoring plan.
What it is: A research grant proposal. It argues for a research programme and reports earlier results, but it is a funding pitch, not a peer-reviewed paper.
Where this came from
- Will Manidis’s “most underrated PDF” thread, July 2023 — routing record: The Most Underrated PDF on the Internet (July 2023)
25 pages. A copy is archived locally against link rot; the header links the original source.