|
|
Unified Querying of Multi-Model Data
Crha, Daniel ; Holubová, Irena (advisor) ; Pokorný, Jaroslav (referee)
The vast majority of current multi-model querying solutions require the user to have intimate knowledge of the specific models involved. There exists a single approach for truly unified multi-model querying, but this approach is not practically usable for most users due to its complexity. In this thesis we present MMQL, a multi-model query lan- guage based on category theory, which was designed using SPARQL as a basis. Using MMQL, users can formulate multi-model, multi-database queries without needing to know about the way the data is stored. We also present our proposal for the implemen- tation of MMQL, including the required supporting algorithms. To verify the validity of our proposal, we built the proof-of-concept tool MM-quecat, an implementation of basic MMQL concepts. We then evaluated MM-quecat in a scenario involving PostgreSQL and MongoDB, querying both databases with a single MMQL query. As we present one of the first ever approaches for unified multi-model querying, we also analyze the weak- nesses and limitations of the proposed approach, opening the door for future iterations and improvements. 1
|
guest ::
