Finish exercise 4 of sheet 3 about gki

[NiklasGebauer]

I'll try b) now.

[cyborgx7]

Ich schau mir den Rest von 4 auch noch an.

[NiklasGebauer]

Finished 4 b).

[cyborgx7]

made small corrections

[NiklasGebauer]

4 c) is done

[cyborgx7]

working on 4d right now. I have an idea, not sure if I'm on the right track though

[NiklasGebauer]

4 e) is finished as well now I guess, but someone should review it before we hand it in.

[cyborgx7]

done with 4d, if somebody could look at it, I'll look at 4e

[NiklasGebauer]

I think the first steps in d) should look like:

1. F(y)
2. Be(x,GKI), Ge(x,Poker) -| (1), {y/x}
   and then you have different ways to fulfill those requirements. You're supposed to choose one possible rule and then use backtracking if you come to a point where you can't match any more requirements to existing rules or facts.

[cyborgx7]

No, you are supposed to use backtracking if you chose a wrong rule, like if I tried to use (3) for Be(S,GKI) eventhough In(S) is not true. But you don't have to branch out for different things left to prove.

[cyborgx7]

How did you choose that specific fact for 4e. I thought you would have to continue with forwardchaining until you can't find anymore facts.

[NiklasGebauer]

You're supposed to use backtracking as soon as you find a requirement in your goals list, that can't be fulfilled by any of the rules or facts given. Don't know if you mean that, but you can check in the lecture's slides. The way you started your proof is still wrong though imo.
You need to take your goal and then 'go backwards' to prove it.
Our goal is to show Fr(y). That's also why I stopped forwardchaining after 3 iterations. It stops (and returns a failure) if you can't produce any more facts OR stops and returns a solution if you find your goal in the new facts created during the last iteration.

[cyborgx7]

du hast recht, hab die frage falsch verstanden. mach das schnell

[cyborgx7]

Like this, or do you think I have to branch for which fact I prove first, as well?

[NiklasGebauer]

I already switched off everything, will have a look at it tomorrow if that's okay for you

-------- Ursprüngliche Nachricht --------
Von: Leonard Witte [email protected]
Datum:
An: TN1ck/GKI [email protected]
Cc: NiklasGebauer [email protected]
Betreff: Re: [GKI] Finish exercise 4 of sheet 3 (#1)

Like this, or do you think I have to branch for which fact I prove first, as well?

—
Reply to this email directly or view it on GitHub.

[cyborgx7]

sure