Algebraic Structure Theory
Theory on the algebraic structuring of databases in Ecology.
The goal of the project is to describe a reasonable algebraic structure that databases should have so that we can faithfully merge disconnected instances of these databases. This problem can arise in ecology and biology in at least two contexts:
- Data collection in the context of field work.
- The accumulation of observations from different sources in a single repository.
For small databases, all inconsistencies and ambiguities could be handled individually, but for large databases, the merge-update operation should be imposed automatically.
Although several techniques have been proposed to solve this and other similar problems which in the literature are referred to as database replication, mobile database reconciliation, or record linkage, much more emphasis has been given now on the fusion speed for very large databases, while accepting a cost that is minimized in the probabilistic sense, sometimes in a very theoretical framework.
Our approach, which is driven by the need for research databases to be error-free, is to impose an algebraic structure on the database schema so that all updates and merges can be guaranteed and faithful to the design.
This algebraic approach also makes it possible to treat databases as dynamic abstract objects with temporal evolution, and to provide higher abstractions that allow new perspectives in data exploration.
The project is jointly led by the Jean Alexandre Dieudonné Laboratory (LJAD) of the Côte d'Azur University and the Monaco Scientific Center (CSM).
Teams involved in the "Algebraic Structure Thoery" Project:
- CSM: Victor PLANAS, responsible for the Artificial Intelligence axis.
- UCA: JA Dieudonné : Frederic PATRAS