AI RESEARCH
MICRO: A Lightweight Middleware for Optimizing Cross-store Cross-model Graph-Relation Joins [Technical Report]
arXiv CS.LG
•
ArXi:2603.13835v1 Announce Type: cross Modern data applications increasingly involve heterogeneous data managed in different models and d across disparate database engines, often deployed as separate installs. Limited research has addressed cross-model query processing in federated environments. This paper takes a step toward bridging this gap by: (1) formally defining a class of cross-model join queries between a graph and a relational by proposing a unified algebra; (2)