Deductive logic is very nice : from OP(A, B) you can produce C. A logical application. That's where the term "application" comes from. You apply an operation OP to inputs A and B and the application produces C which is the result you want.
Inductive logic is more difficult : you have C and OP and you want to find A and B. It's like going backwards or up in the usual falling cascade of deductions that our brain and programs produce. Usually to find A and B people have some sort of caching memory ─ and a pattern-matcher. They already know about the deductive relation and can produce induction by being exhaustive on prepared knowledge and by elimination. Actually this is still a deductive process as our brain is a directed graph and application is all we learn at school. To become an inductive persona you need to make loops in your brain. Loop the result to the premices and remember. Next time you'll know : this is a rotation, this is a translation, this a scaling with factor 4.2. You have to know it to actually see it and name it.
When you think about models theory it gives more insight :
Everything is Subject, Predicate, Object. Then deduction becomes (S, P) → O. That was the main argument for Prolog is that people know about deduction. Where prolog shines is that it can do induction just as well as deduction because the (deductive) operation becomes (P, O) → S and you can reason (pattern match) on all the possibilities.
So a graph database with SPO, POS, OSP indexes allows you to do induction and even predicate recognition. You can also imagine (O, S) → P which is the formal way to do analysis.
In a graph database all data flows and can become either a Subject (which they call an object in object oriented programming), a Predicate (why not call it a method if it's functional) and an Object which is the result of applying the predicate to the subject which they call a value most of the time in programming.
In KC3 the graph database is just that : three indexes and it allows for
And it works on everything you put in the database.