![A simple reflex agent
• Rules to map percepts into observations:
b,g,u,c,t Percept([Stench, b, g, u, c], t) Stench(t)
s,g,u,c,t Percept([s, Breeze, g, u, c], t) Breeze(t)
s,b,u,c,t Percept([s, b, Glitter, u, c], t) AtGold(t)
• Rules to select an action given observations:
t AtGold(t) Action(Grab, t);
• Some difficulties:
– Consider Climb. There is no percept that indicates the agent should
climb out – position and holding gold are not part of the percept
sequence
– Loops – the percept will be repeated when you return to a square,
which should cause the same response (unless we maintain some
internal model of the world)](https://image.slidesharecdn.com/firstorderlogic-240719033749-4e70e991/85/Knowledge-Representation-with-predicate-knoledgeFirstOrderLogic-ppt-42-320.jpg)
![The frame problem II
• Successor-state axiom: General statement that
characterizes every way in which a particular predicate can
become true:
– Either it can be made true, or it can already be true and not be
changed:
– On (x, table, Result(a,s))
[On (x, z, s) Clear (x, s) a = Move(x, table)]
[On (x, table, s) a Move (x, z)]
• In complex worlds, where you want to reason about longer
chains of action, even these types of axioms are too
cumbersome
– Planning systems use special-purpose inference methods to reason
about the expected state of the world at any point in time during a
multi-step plan](https://image.slidesharecdn.com/firstorderlogic-240719033749-4e70e991/85/Knowledge-Representation-with-predicate-knoledgeFirstOrderLogic-ppt-49-320.jpg)
![Goal-based agents
• Once the gold is found, it is necessary to change strategies.
So now we need a new set of action values.
• We could encode this as a rule:
– (s) Holding(Gold,s) GoalLocation([1,1]),s)
• We must now decide how the agent will work out a
sequence of actions to accomplish the goal.
• Three possible approaches are:
– Inference: good versus wasteful solutions
– Search: make a problem with operators and set of states
– Planning: to be discussed later](https://image.slidesharecdn.com/firstorderlogic-240719033749-4e70e991/85/Knowledge-Representation-with-predicate-knoledgeFirstOrderLogic-ppt-56-320.jpg)
Knowledge Representation with predicate knoledgeFirstOrderLogic.ppt
- 2. Outline • First-order logic – Properties, relations, functions, quantifiers, … – Terms, sentences, axioms, theories, proofs, … • Extensions to first-order logic • Logical agents – Reflex agents – Representing change: situation calculus, frame problem – Preferences on actions – Goal-based agents
- 4. Predicate Logic • Predicate logic is an extension of propositional logic. • A= The Ball’s color is Red ( propositional) • Color(Ball,Red) Predicate is Color and Ball and Red are argument. • Rohan Likes Banana • Likes(Rohan, Banana) • Where likes is a predicate and rohan and banana are argument
- 5. Predicate Logic • Quantifies : We need include all and some • Universal ∀ • Existential ∃ • Everybody loves somebody • ∀x∃y L(x,y)=loves(everybody,somebody)
- 6. First-order logic • First-order logic (FOL) models the world in terms of – Objects, which are things with individual identities – Properties of objects that distinguish them from other objects – Relations that hold among sets of objects – Functions, which are a subset of relations where there is only one “value” for any given “input” • Examples: – Objects: Students, lectures, companies, cars ... – Relations: Brother-of, bigger-than, outside, part-of, has-color, occurs-after, owns, visits, precedes, ... – Properties: blue, oval, even, large, ... – Functions: father-of, best-friend, second-half, one-more-than ...
- 10. User provides • Constant symbols, which represent individuals in the world – Mary – 3 – Green • Function symbols, which map individuals to individuals – father-of(Mary) = John – color-of(Sky) = Blue • Predicate symbols, which map individuals to truth values – greater(5,3) – green(Grass) – color(Grass, Green)
- 11. FOL Provides • Variable symbols – E.g., x, y, foo • Connectives – Same as in PL: not (), and (), or (), implies (), if and only if (biconditional ) • Quantifiers – Universal x or (Ax) – Existential x or (Ex)
- 12. Sentences are built from terms and atoms • A term (denoting a real-world individual) is a constant symbol, a variable symbol, or an n-place function of n terms. x and f(x1, ..., xn) are terms, where each xi is a term. A term with no variables is a ground term • An atomic sentence (which has value true or false) is an n-place predicate of n terms • A complex sentence is formed from atomic sentences connected by the logical connectives: P, PQ, PQ, PQ, PQ where P and Q are sentences • A quantified sentence adds quantifiers and • A well-formed formula (wff) is a sentence containing no “free” variables. That is, all variables are “bound” by universal or existential quantifiers. (x)P(x,y) has x bound as a universally quantified variable, but y is free.
- 13. Quantifiers • Universal quantification – (x)P(x) means that P holds for all values of x in the domain associated with that variable – E.g., (x) dolphin(x) mammal(x) • Existential quantification – ( x)P(x) means that P holds for some value of x in the domain associated with that variable – E.g., ( x) mammal(x) lays-eggs(x) – Permits one to make a statement about some object without naming it
- 14. Quantifiers • Universal quantifiers are often used with “implies” to form “rules”: (x) student(x) smart(x) means “All students are smart” • Universal quantification is rarely used to make blanket statements about every individual in the world: (x)student(x)smart(x) means “Everyone in the world is a student and is smart” • Existential quantifiers are usually used with “and” to specify a list of properties about an individual: (x) student(x) smart(x) means “There is a student who is smart” • A common mistake is to represent this English sentence as the FOL sentence: (x) student(x) smart(x) – But what happens when there is a person who is not a student?
- 15. Quantifier Scope • Switching the order of universal quantifiers does not change the meaning: – (x)(y)P(x,y) ↔ (y)(x) P(x,y) • Similarly, you can switch the order of existential quantifiers: – (x)(y)P(x,y) ↔ (y)(x) P(x,y) • Switching the order of universals and existentials does change meaning: – Everyone likes someone: (x)(y) likes(x,y) – Someone is liked by everyone: (y)(x) likes(x,y)
- 16. Connections between All and Exists We can relate sentences involving and using De Morgan’s laws: (x) P(x) ↔ (x) P(x) (x) P ↔ (x) P(x) (x) P(x) ↔ (x) P(x) (x) P(x) ↔ (x) P(x)
- 17. Quantified inference rules • Universal instantiation – x P(x) P(A) • Universal generalization – P(A) P(B) … x P(x) • Existential instantiation – x P(x) P(F) skolem constant F • Existential generalization – P(A) x P(x)
- 18. Universal instantiation (a.k.a. universal elimination) • If (x) P(x) is true, then P(C) is true, where C is any constant in the domain of x • Example: (x) eats(Ziggy, x) eats(Ziggy, IceCream) • The variable symbol can be replaced by any ground term, i.e., any constant symbol or function symbol applied to ground terms only
- 19. Existential instantiation (a.k.a. existential elimination) • From (x) P(x) infer P(c) • Example: – (x) eats(Ziggy, x) eats(Ziggy, Stuff) • Note that the variable is replaced by a brand-new constant not occurring in this or any other sentence in the KB • Also known as skolemization; constant is a skolem constant • In other words, we don’t want to accidentally draw other inferences about it by introducing the constant • Convenient to use this to reason about the unknown object, rather than constantly manipulating the existential quantifier
- 20. Existential generalization (a.k.a. existential introduction) • If P(c) is true, then (x) P(x) is inferred. • Example eats(Ziggy, IceCream) (x) eats(Ziggy, x) • All instances of the given constant symbol are replaced by the new variable symbol • Note that the variable symbol cannot already exist anywhere in the expression
- 21. Translating English to FOL Every gardener likes the sun. x gardener(x) likes(x,Sun) You can fool some of the people all of the time. x t person(x) time(t) can-fool(x,t) You can fool all of the people some of the time. x t (person(x) time(t) can-fool(x,t)) x (person(x) t (time(t) can-fool(x,t)) All purple mushrooms are poisonous. x (mushroom(x) purple(x)) poisonous(x) No purple mushroom is poisonous. x purple(x) mushroom(x) poisonous(x) x (mushroom(x) purple(x)) poisonous(x) There are exactly two purple mushrooms. x y mushroom(x) purple(x) mushroom(y) purple(y) ^ (x=y) z (mushroom(z) purple(z)) ((x=z) (y=z)) Clinton is not tall. tall(Clinton) X is above Y iff X is on directly on top of Y or there is a pile of one or more other objects directly on top of one another starting with X and ending with Y. x y above(x,y) ↔ (on(x,y) z (on(x,z) above(z,y))) Equivalent Equivalent
- 22. Monty Python and The Art of Fallacy Cast –Sir Bedevere the Wise, master of (odd) logic –King Arthur –Villager 1, witch-hunter –Villager 2, ex-newt –Villager 3, one-line wonder –All, the rest of you scoundrels, mongrels, and nere-do-wells.
- 23. An example from Monty Python by way of Russell & Norvig • FIRST VILLAGER: We have found a witch. May we burn her? • ALL: A witch! Burn her! • BEDEVERE: Why do you think she is a witch? • SECOND VILLAGER: She turned me into a newt. • B: A newt? • V2 (after looking at himself for some time): I got better. • ALL: Burn her anyway. • B: Quiet! Quiet! There are ways of telling whether she is a witch.
- 24. Monty Python cont. • B: Tell me… what do you do with witches? • ALL: Burn them! • B: And what do you burn, apart from witches? • Third Villager: …wood? • B: So why do witches burn? • V2 (after a beat): because they’re made of wood? • B: Good. • ALL: I see. Yes, of course.
- 25. Monty Python cont. • B: So how can we tell if she is made of wood? • V1: Make a bridge out of her. • B: Ah… but can you not also make bridges out of stone? • ALL: Yes, of course… um… er… • B: Does wood sink in water? • ALL: No, no, it floats. Throw her in the pond. • B: Wait. Wait… tell me, what also floats on water? • ALL: Bread? No, no no. Apples… gravy… very small rocks… • B: No, no, no,
- 26. Monty Python cont. • KING ARTHUR: A duck! • (They all turn and look at Arthur. Bedevere looks up, very impressed.) • B: Exactly. So… logically… • V1 (beginning to pick up the thread): If she… weighs the same as a duck… she’s made of wood. • B: And therefore? • ALL: A witch!
- 27. Monty Python Fallacy #1 • x witch(x) burns(x) • x wood(x) burns(x) • ------------------------------- • z witch(x) wood(x) • p q • r q • --------- • p r Fallacy: Affirming the conclusion
- 28. Monty Python Near-Fallacy #2 • wood(x) can-build-bridge(x) • ----------------------------------------- • can-build-bridge(x) wood(x) • B: Ah… but can you not also make bridges out of stone?
- 29. Monty Python Fallacy #3 • x wood(x) floats(x) • x duck-weight (x) floats(x) • ------------------------------- • x duck-weight(x) wood(x) • p q • r q • ----------- • r p
- 30. Monty Python Fallacy #4 • z light(z) wood(z) • light(W) • ------------------------------ • wood(W) ok………….. • witch(W) wood(W) applying universal instan. to fallacious conclusion #1 • wood(W) • --------------------------------- • witch(z)
- 31. Example: A simple genealogy KB by FOL • Build a small genealogy knowledge base using FOL that – contains facts of immediate family relations (spouses, parents, etc.) – contains definitions of more complex relations (ancestors, relatives) – is able to answer queries about relationships between people • Predicates: – parent(x, y), child(x, y), father(x, y), daughter(x, y), etc. – spouse(x, y), husband(x, y), wife(x,y) – ancestor(x, y), descendant(x, y) – male(x), female(y) – relative(x, y) • Facts: – husband(Joe, Mary), son(Fred, Joe) – spouse(John, Nancy), male(John), son(Mark, Nancy) – father(Jack, Nancy), daughter(Linda, Jack) – daughter(Liz, Linda) – etc.
- 32. • Rules for genealogical relations – (x,y) parent(x, y) ↔ child (y, x) (x,y) father(x, y) ↔ parent(x, y) male(x) (similarly for mother(x, y)) (x,y) daughter(x, y) ↔ child(x, y) female(x) (similarly for son(x, y)) – (x,y) husband(x, y) ↔ spouse(x, y) male(x) (similarly for wife(x, y)) (x,y) spouse(x, y) ↔ spouse(y, x) (spouse relation is symmetric) – (x,y) parent(x, y) ancestor(x, y) (x,y)(z) parent(x, z) ancestor(z, y) ancestor(x, y) – (x,y) descendant(x, y) ↔ ancestor(y, x) – (x,y)(z) ancestor(z, x) ancestor(z, y) relative(x, y) (related by common ancestry) (x,y) spouse(x, y) relative(x, y) (related by marriage) (x,y)(z) relative(z, x) relative(z, y) relative(x, y) (transitive) (x,y) relative(x, y) ↔ relative(y, x) (symmetric) • Queries – ancestor(Jack, Fred) /* the answer is yes */ – relative(Liz, Joe) /* the answer is yes */ – relative(Nancy, Matthew) /* no answer in general, no if under closed world assumption */ – (z) ancestor(z, Fred) ancestor(z, Liz)
- 33. Semantics of FOL • Domain M: the set of all objects in the world (of interest) • Interpretation I: includes – Assign each constant to an object in M – Define each function of n arguments as a mapping Mn => M – Define each predicate of n arguments as a mapping Mn => {T, F} – Therefore, every ground predicate with any instantiation will have a truth value – In general there is an infinite number of interpretations because |M| is infinite • Define logical connectives: ~, ^, , =>, <=> as in PL • Define semantics of (x) and (x) – (x) P(x) is true iff P(x) is true under all interpretations – (x) P(x) is true iff P(x) is true under some interpretation
- 34. • Model: an interpretation of a set of sentences such that every sentence is True • A sentence is – satisfiable if it is true under some interpretation – valid if it is true under all possible interpretations – inconsistent if there does not exist any interpretation under which the sentence is true • Logical consequence: S |= X if all models of S are also models of X
- 35. Axioms, definitions and theorems •Axioms are facts and rules that attempt to capture all of the (important) facts and concepts about a domain; axioms can be used to prove theorems –Mathematicians don’t want any unnecessary (dependent) axioms –ones that can be derived from other axioms –Dependent axioms can make reasoning faster, however –Choosing a good set of axioms for a domain is a kind of design problem •A definition of a predicate is of the form “p(X) ↔ …” and can be decomposed into two parts –Necessary description: “p(x) …” –Sufficient description “p(x) …” –Some concepts don’t have complete definitions (e.g., person(x))
- 36. More on definitions • A necessary condition must be satisfied for a statement to be true. • A sufficient condition, if satisfied, assures the statement’s truth. • Duality: “P is sufficient for Q” is the same as “Q is necessary for P.” • Examples: define father(x, y) by parent(x, y) and male(x) – parent(x, y) is a necessary (but not sufficient) description of father(x, y) • father(x, y) parent(x, y) – parent(x, y) ^ male(x) ^ age(x, 35) is a sufficient (but not necessary) description of father(x, y): father(x, y) parent(x, y) ^ male(x) ^ age(x, 35) – parent(x, y) ^ male(x) is a necessary and sufficient description of father(x, y) parent(x, y) ^ male(x) ↔ father(x, y)
- 37. More on definitions P(x) S(x) S(x) is a necessary condition of P(x) (x) P(x) => S(x) S(x) P(x) S(x) is a sufficient condition of P(x) (x) P(x) <= S(x) P(x) S(x) S(x) is a necessary and sufficient condition of P(x) (x) P(x) <=> S(x)
- 38. Higher-order logic • FOL only allows to quantify over variables, and variables can only range over objects. • HOL allows us to quantify over relations • Example: (quantify over functions) “two functions are equal iff they produce the same value for all arguments” f g (f = g) (x f(x) = g(x)) • Example: (quantify over predicates) r transitive( r ) (xyz) r(x,y) r(y,z) r(x,z)) • More expressive, but undecidable. (there isn’t an effective algorithm to decide whether all sentences are valid) – First-order logic is decidable only when it uses predicates with only one argument.
- 39. Expressing uniqueness • Sometimes we want to say that there is a single, unique object that satisfies a certain condition • “There exists a unique x such that king(x) is true” – x king(x) y (king(y) x=y) – x king(x) y (king(y) xy) – ! x king(x) • “Every country has exactly one ruler” – c country(c) ! r ruler(c,r) • Iota operator: “ x P(x)” means “the unique x such that p(x) is true” – “The unique ruler of Freedonia is dead” – dead( x ruler(freedonia,x))
- 40. Notational differences • Different symbols for and, or, not, implies, ... – – p v (q ^ r) – p + (q * r) – etc • Prolog cat(X) :- furry(X), meows (X), has(X, claws) • Lispy notations (forall ?x (implies (and (furry ?x) (meows ?x) (has ?x claws)) (cat ?x)))
- 41. Logical agents for the Wumpus World Three (non-exclusive) agent architectures: –Reflex agents • Have rules that classify situations, specifying how to react to each possible situation –Model-based agents • Construct an internal model of their world –Goal-based agents • Form goals and try to achieve them
- 42. A simple reflex agent • Rules to map percepts into observations: b,g,u,c,t Percept([Stench, b, g, u, c], t) Stench(t) s,g,u,c,t Percept([s, Breeze, g, u, c], t) Breeze(t) s,b,u,c,t Percept([s, b, Glitter, u, c], t) AtGold(t) • Rules to select an action given observations: t AtGold(t) Action(Grab, t); • Some difficulties: – Consider Climb. There is no percept that indicates the agent should climb out – position and holding gold are not part of the percept sequence – Loops – the percept will be repeated when you return to a square, which should cause the same response (unless we maintain some internal model of the world)
- 43. Representing change • Representing change in the world in logic can be tricky. • One way is just to change the KB – Add and delete sentences from the KB to reflect changes – How do we remember the past, or reason about changes? • Situation calculus is another way • A situation is a snapshot of the world at some instant in time • When the agent performs an action A in situation S1, the result is a new situation S2.
- 44. Situations
- 45. Situation calculus • A situation is a snapshot of the world at an interval of time during which nothing changes • Every true or false statement is made with respect to a particular situation. – Add situation variables to every predicate. – at(Agent,1,1) becomes at(Agent,1,1,s0): at(Agent,1,1) is true in situation (i.e., state) s0. – Alternatively, add a special 2nd-order predicate, holds(f,s), that means “f is true in situation s.” E.g., holds(at(Agent,1,1),s0) • Add a new function, result(a,s), that maps a situation s into a new situation as a result of performing action a. For example, result(forward, s) is a function that returns the successor state (situation) to s • Example: The action agent-walks-to-location-y could be represented by – (x)(y)(s) (at(Agent,x,s) onbox(s)) at(Agent,y,result(walk(y),s))
- 46. Deducing hidden properties • From the perceptual information we obtain in situations, we can infer properties of locations l,s at(Agent,l,s) Breeze(s) Breezy(l) l,s at(Agent,l,s) Stench(s) Smelly(l) • Neither Breezy nor Smelly need situation arguments because pits and Wumpuses do not move around
- 47. Deducing hidden properties II • We need to write some rules that relate various aspects of a single world state (as opposed to across states) • There are two main kinds of such rules: – Causal rules reflect the assumed direction of causality in the world: (l1,l2,s) At(Wumpus,l1,s) Adjacent(l1,l2) Smelly(l2) ( l1,l2,s) At(Pit,l1,s) Adjacent(l1,l2) Breezy(l2) Systems that reason with causal rules are called model-based reasoning systems – Diagnostic rules infer the presence of hidden properties directly from the percept-derived information. We have already seen two diagnostic rules: ( l,s) At(Agent,l,s) Breeze(s) Breezy(l) ( l,s) At(Agent,l,s) Stench(s) Smelly(l)
- 48. Representing change: The frame problem • Frame axioms: If property x doesn’t change as a result of applying action a in state s, then it stays the same. – On (x, z, s) Clear (x, s) On (x, table, Result(Move(x, table), s)) On(x, z, Result (Move (x, table), s)) – On (y, z, s) y x On (y, z, Result (Move (x, table), s)) – The proliferation of frame axioms becomes very cumbersome in complex domains
- 49. The frame problem II • Successor-state axiom: General statement that characterizes every way in which a particular predicate can become true: – Either it can be made true, or it can already be true and not be changed: – On (x, table, Result(a,s)) [On (x, z, s) Clear (x, s) a = Move(x, table)] [On (x, table, s) a Move (x, z)] • In complex worlds, where you want to reason about longer chains of action, even these types of axioms are too cumbersome – Planning systems use special-purpose inference methods to reason about the expected state of the world at any point in time during a multi-step plan
- 50. Qualification problem • Qualification problem: – How can you possibly characterize every single effect of an action, or every single exception that might occur? – When I put my bread into the toaster, and push the button, it will become toasted after two minutes, unless… • The toaster is broken, or… • The power is out, or… • I blow a fuse, or… • A neutron bomb explodes nearby and fries all electrical components, or… • A meteor strikes the earth, and the world we know it ceases to exist, or…
- 51. Ramification problem • Similarly, it’s just about impossible to characterize every side effect of every action, at every possible level of detail: – When I put my bread into the toaster, and push the button, the bread will become toasted after two minutes, and… • The crumbs that fall off the bread onto the bottom of the toaster over tray will also become toasted, and… • Some of the aforementioned crumbs will become burnt, and… • The outside molecules of the bread will become “toasted,” and… • The inside molecules of the bread will remain more “breadlike,” and… • The toasting process will release a small amount of humidity into the air because of evaporation, and… • The heating elements will become a tiny fraction more likely to burn out the next time I use the toaster, and… • The electricity meter in the house will move up slightly, and…
- 52. Knowledge engineering! • Modeling the “right” conditions and the “right” effects at the “right” level of abstraction is very difficult • Knowledge engineering (creating and maintaining knowledge bases for intelligent reasoning) is an entire field of investigation • Many researchers hope that automated knowledge acquisition and machine learning tools can fill the gap: – Our intelligent systems should be able to learn about the conditions and effects, just like we do! – Our intelligent systems should be able to learn when to pay attention to, or reason about, certain aspects of processes, depending on the context!
- 53. Preferences among actions • A problem with the Wumpus world knowledge base that we have built so far is that it is difficult to decide which action is best among a number of possibilities. • For example, to decide between a forward and a grab, axioms describing when it is OK to move to a square would have to mention glitter. • This is not modular! • We can solve this problem by separating facts about actions from facts about goals. This way our agent can be reprogrammed just by asking it to achieve different goals.
- 54. Preferences among actions • The first step is to describe the desirability of actions independent of each other. • In doing this we will use a simple scale: actions can be Great, Good, Medium, Risky, or Deadly. • Obviously, the agent should always do the best action it can find: (a,s) Great(a,s) Action(a,s) (a,s) Good(a,s) (b) Great(b,s) Action(a,s) (a,s) Medium(a,s) ((b) Great(b,s) Good(b,s)) Action(a,s) ...
- 55. Preferences among actions • We use this action quality scale in the following way. • Until it finds the gold, the basic strategy for our agent is: – Great actions include picking up the gold when found and climbing out of the cave with the gold. – Good actions include moving to a square that’s OK and hasn't been visited yet. – Medium actions include moving to a square that is OK and has already been visited. – Risky actions include moving to a square that is not known to be deadly or OK. – Deadly actions are moving into a square that is known to have a pit or a Wumpus.
- 56. Goal-based agents • Once the gold is found, it is necessary to change strategies. So now we need a new set of action values. • We could encode this as a rule: – (s) Holding(Gold,s) GoalLocation([1,1]),s) • We must now decide how the agent will work out a sequence of actions to accomplish the goal. • Three possible approaches are: – Inference: good versus wasteful solutions – Search: make a problem with operators and set of states – Planning: to be discussed later