Sparse Matrix-Based Substitution for Cross-Model Graph Database Mappings (Master Project, Ongoing)
Author
Description
Graphs are foundational structures across many disciplines, from theoretical mathematics to the core design of computer networks. Within data management, graph databases represent a unique category. While they function as data stores, their inherent structure, encoding relationships and leveraging specialized, graph-centric tools, suggests they contain a sophisticated logic layer that goes beyond the traditional responsibilities of a pure database system.
This distinction has driven interest in substitution approaches: replicating graph functionality by adding an application layer on top of existing non-graph databases (e.g., document or relational stores). The rising popularity of multi-model databases has amplified the need for capable, performant substitution methods, particularly as the storage of graph data is relatively straightforward to map.
However, the logic-based retrieval of graph structures and the execution of graph operations on non-graph stores have received comparatively limited research attention. This is the critical gap this project addresses.
This project lays the groundwork to bridge this gap by exploring the use of sparse matrix-based substitutions to perform complex graph operations on data stored within non-graph database systems. We hypothesize that by representing graph structure as a sparse matrix, we can leverage established matrix operations to efficiently execute graph algorithms.
Start / End Dates
2025/11/17 - 2025/05/17