The semantic web, which had a kind of graph database (Graph DB) — triplestores — and its related resource description framework (RDF) as one of its core part, generated many expectations. For instance, it was expected that triplestores would replace relational database management systems (RDBMS). Similarly, it was anticipated RDF would displace other exchange formats. In addition, graph query languages such as the protocol and RDF query language (SPARQL) would become natural and simple to understand.
However, Graph DBs failed to stand up to the test of real enterprise use cases in certain aspects. For example, Graph DBs fell short as a panacea for data model changes because of the modifications required in loading and application logic. The hope that a graph representation like RDF would be convenient for anything also fell short of expectations. The expectations were belied because RDFs were excruciating to use in the real world to express metadata concepts such as time/provenance/trust due to their simplicity (triples). Furthermore, the premise that graph DB query languages would be superior in all use cases turned out to be a matter of conjecture. In real-world scenarios, some graph query languages like SPARQL were natural for certain questions and contrived for others.
Today, Graph DBs are describing an upward curve for several valid reasons. To begin, flexible and enterprise-ready Graph DB engines have emerged. Furthermore, enterprise adoption has grown because, in contrast to the experimental approach taken earlier, Graph DBs are currently leveraged intelligently for use cases where they offer critical benefits. In general, Graph DBs are unbeatable for exploring in-depth relationships between records. They are efficient in detecting suspect rings to fight fraud, determining the shortest path or other path related algorithms in law enforcement and cybersecurity, and computing graphic metrics.
In conclusion, in spite of the earlier drawbacks, today there is a compelling case for Graph DBs as part of enterprise knowledge graphs.
Click here to read the article written by Giovanni Tummarello, founder and chief product officer at Siren.
Please give your feedback on this article or share a similar story for publishing by clicking here.
