It is not feasible to store transactional data in ontologies. when the size of data will grow, the number of triples in the ontology will be in million or trillion. Due to the un-structured file structure of ontologies, it is not possible to traverse ontologies with trillions of tripples even with the help of the latest fastest processor and un-limited memory (primary, cache, registers, etc).
what do you say? waiting for your valuable comments.
Hi Abdul Sattar
maybe you could change your mind and switch from your original idea:
- to "store transactional data in ontology"
Conference Paper OntoDB: An Ontology-Based Database for Data Intensive Applications
- to "store ontology in database"
Article The relational database layout to store ontology knowledge base
If I think about what you would like to achieve without an analysis and study or overview of the domain (and the project), I would store the transactional data in a JSON structure as a "knowledge base" for each transaction (litteraly, each transaction is a knowledge graph about this transaction and its context etc.) which follow the main ontology and store them in a noSQL database for example.
After you could request the transactional database based on the ontology and generate more global knowledge graphs and make operations on
|_ (Data) all the "transactional data":
|___ (Information) to exact "transactional information" ,
|______ (Knowledge) to generate new "transactional knowledge",
|________ (Wisdom/Intelligence) and why not, based on the "transactional intelligence & wisdom" produced,
|___________ trigger proactive or preemptive actions by an Ethical and Consequence Engine
It seems that my idea is closed with the idea of this proposal:
Conference Paper OntoDB: An Ontology-Based Database for Data Intensive Applications
"Recently, several approaches and systems were proposed to store in the same database data and the ontologies describing their meanings. We call these databases, ontology-based databases (OBDBs), and Ontology-based data denotes those data that represent ontology individuals (i.e., instance of ontology classes). To ensure a high performance of queries on top of these OBDBs, efficient representation of ontology-based data becomes a new challenge. [...]"
https://www.lias-lab.fr/publications/7690/2007-ICDSAA-Dehainsala.pdf
Laurent Berry Thanks for your detailed and comprehensive answer. I liked the ontology-based databases. This hybrid approach will assist in solving most of the issues related to storage and query performances. Thanks again.
Abdul Sattar ,
you should investigate and test Neo4J https://neo4j.com/ which is a commercial solution but really interesting to manage data and their relations in a database. There are a good demos and tutorials on their website: https://neo4j.com/developer/
Maybe they offer licences for researchers (I'll test for me in December)
"Neo4j is a native graph database, built from the ground up to leverage not only data but also data relationships. Neo4j connects data as it’s stored, enabling queries never before imagined, at speeds never thought possible."
- Badges
- Science topic
- Similar topics
- Comment