GraphDB

Kommentare: There are other triple stores with different feature sets, but it don't think there's any triplestore that is better than GraphDB.

Vorteile: It's very quick and easy to deploy GraphDB, ingest some RDF data, build queries in a IDE-like environment, and visualize relationships. A semantic similarity search tool is provided.

Nachteile: The OntoRefine tool is great for converting tabular data files into semantic triples, but there's no support for reading from relational databases. There are nice free text indexing & search tools, but no natural language parser for discovering entities and relationships. There are several pre-configured reasoning levels plus support for writing one's own rules, but no support for SWRL. Like most triplesotres, OWL2 reasoning over complex axioms and millions of data triples isn't fun/fast/realistic? (I say that based on a single node, two threads, and 256 GB RAM.)

Kommentare: We use GraphDB together with PoolParty as part of the Semantic Integrator solution.
We use GraphDB as our test triple store.
We use GraphDB to publish our "small" linked open data sets.

Vorteile: I can be very short about my (our) experiences so far with GraphDB.
GraphDB is a clear winner for our usecases now. The learning curve is not steep, almost self-explanatory. It’s fast and it fits our needs; for now. We loved the graphics of GraphDB.

Nachteile: OntoRefine. It looks fine but we missed some modeling features. We switched back to OpenRefine.

Kommentare: We've had a really good time working with it. Once we have a good model definition (which applies for any graph engine) it simply works and works really well.

Vorteile: Over the last years, GraphDB has improved a lot, speaking about performance and inference. Graph DBs have always had the problem of being too slow for solving queries, we always had to structure them in a way that it was optimal for the engine to solve them. Graph DB has been always the most performant one.
Another feature that I loved from it is that, in order to install it, you simply copy a single jar file and you're almost ready to go. The user interface helps a lot. And its compliance with several standards for the interchange formats makes the way pretty straight forward no matter what tool was used to generate an interchange file.

Nachteile: The con I find with the product is about updates. When a new version comes up (which sometimes I don't know about the fact that there's a new release), I need to manually download it and deploy it. Also it would be desirable if we had access to connectors for different platforms. We work a lot with node.js and there's almost no libraries to use it with graphDB that are backed by ontotext.

Kommentare: The use of graphDB solves many speed issues as it is very fast.
The webservice api allows manipulation of the repositories (create, delete, import).
The only problem is that after some updates, the speed of the endpoint queries dorps very much.
An example for this if the migration from v8.11.1 to v9.0.
I am using the free version but in a single thread and they mention that the only limitation of the free version is the limitation of simultaneous queries to only two.
I had the same problems when i tried the enterprise edition for a trial period and never managed to reach the speed of the free version with any change I tried in the setup or options.

Vorteile: Sparql support.
Ease of setup.
Webservice api.
Functionality of free version.

Nachteile: Speed changes after some updates.
Removal from GCC

Kommentare: We are currently using graphdb in a PoC as semantic web stack compliant database for data integration in a laboratory environment.

Vorteile: - ease of use (compared to other semantic web stack solutions)
- degree of inferencing implementation
- solution for transforming relational data into RDF with OpenRefine integration
- query performance (for SELECT, a evaluation for INSERT queries could not be given due to use of free version)
- good support even at free version

Nachteile: - versioning of data (see changes over time)
- better controllability of role and rights (give rights for specific graphs in repository)
- no IdP based authentiaction like OpenID Connect (or something based on oAuth2 or at least SAML)
- easy to use integration for object storage (like AWS S3)
- documentation could be more detailed in some places