RDF, Triples, and the Semantic Web in everyday applications

Question

Vague, uninformed questions:

1: Why do almost 100% of application developers, application and literature developer communities (books, tutorials, etc.) assume that you want to express data using a relational database or key store?

2: Why don't everyone use Triple data structures?

3: Are Triples applicable to every problem that relational databases and key value stores are, and are Triples at least as easy to use?

Answer 1

Triples can represent any other data structure. But this does not necessarily facilitate their work. If your problem is tabular, the data table structure will work better. Using the graph data structure, you need to think about how to make tables of triples, and that extra work.

The solution to most problems (especially simple problems when the shape of your data is predictable) does not require the flexibility of the graph data structure.

Answer 2

• Most developers use relational databases and / or key / value stores because they are well-known, widely taught, easily accessible, and adequate for most of what most developers care about.
• Most developers see a small (if any) reason for using triples, except (possibly) for several special purposes (and even the latter is somewhat unusual).
• No - triples are not particularly convenient to use when most people do not understand them and how to use them. Even those developers who understand them usually don't care about what they provide anyway.

More generally, I think that quite a few developers are quickly lost in the confusing mess of RDF, OWL, SKOS, ontologies, reasoning machines, etc. For someone who thinks, “but I just want the user's order history,” (or something else), all this is too much to take part, deal with, etc.

Answer 3

I asked myself the same question before. Typically, people call the complexity of the problem. This is really a bad habit, because the longer we leave the problem, the worse it will be. Semantic Web is a complex solution to a complex problem. It does not get any easier. I also think that comparing simplicity with RDBMS is naive. Most developers today are familiar with ORM and work with abstract tenacity; some never know about the preservation mechanism. Persistence structures for the Semantic Web (ORDFM) are usually not complex or advanced. Having said that, many organizations leave the DBMS and invest in NoSQL solutions, for which RDF and SPARQL, in my opinion, are the best candidates.

An excellent case study that I always point out when people talk about the complexity of the Semantic Web is the story of Bart van Leuven:

http://semtechbizsf2012.semanticweb.com/sessionPop.cfm?confid=65&proposalid=4590

If a genuine full-fledged fireman (who makes real fires) can use SPARQL and RDF instead of databases and proprietary formats to solve a real problem (data availability in the emergency service), then there is little excuse for the rest of us not to. I want to say that this is not a technology that is a barrier, it is something else.