SlideShare a Scribd company logo
1 of 144
Download to read offline
Argumentation for Agent Societies
             Part I

4%33(5()(%$/&6(%7/5($&         !"#"$%&'())%*%&
 8$(9"#3(*%:&;(&<#"37(%&   +,-+.&!/01(%&.$20/)(3&


                                                    1
Introduction to the tutorial


Argumentation for Agent Societies

    (Some) answers to the following two questions:
    1.  What’s argumentation?
            ! mainly today
    2.    What is argumentation good for (in the MAS context)?
            ! mainly tomorrow



   Let’s start with the first question…




                                                                 2
An informal example (1)

   We should run Large Hadron Collider               The conclusion


  LHC allows us to        Understanding
understand the Laws      the Laws of the             The reason
   of the Universe       Universe is good



        We are justified in believing that we should run LHC   !




                                                                      3
An informal example (2)

   We should run Large Hadron Collider                  The conclusion


  LHC allows us to        Understanding
understand the Laws      the Laws of the                The reason
   of the Universe       Universe is good



        We are justified in believing that we should run LHC   !
                                 BUT

       In Argumentation (and in real life as well):
           - reasons are not necessary “conclusive”
            (they don’t logically entail conclusions)
           - arguments and conclusions can be “retracted”
            in front of new information, i.e. counterarguments
                                                                         4
An informal example (3)

   We should run Large Hadron Collider           We should not run LHC



  LHC allows us to        Understanding     LHC will generate     Destroying
understand the Laws      the Laws of the      black holes           Earth
   of the Universe       Universe is good   destroying Earth        is bad



   Now we are justified in believing that we should not run LHC   "




                                                                               5
An informal example (4)

   We should run Large Hadron Collider             We should not run LHC



  LHC allows us to        Understanding      LHC will generate     Destroying
understand the Laws      the Laws of the       black holes           Earth
   of the Universe       Universe is good    destroying Earth        is bad




                                              Black holes will
                                             not destroy Earth

                                               Black holes will
                                             evaporate because
                                            of Hawking radiation




  Now we are again justified in believing that we should run LHC     !
                                                                                6
An informal example (5)

   We should run Large Hadron Collider              We should not run LHC



  LHC allows us to         Understanding      LHC will generate     Destroying
understand the Laws       the Laws of the       black holes           Earth
   of the Universe        Universe is good    destroying Earth        is bad




      Hawking radiation                        Black holes will
       does not exist                         not destroy Earth

                                                Black holes will
     Dr Azzeccagarbugli
                                              evaporate because
          says so
                                             of Hawking radiation



 Now we are again justified in believing that we should not run LHC         "
                                                                                 7
An informal example (6)

   We should run Large Hadron Collider                We should not run LHC



  LHC allows us to          Understanding        LHC will generate    Destroying
understand the Laws        the Laws of the         black holes          Earth
   of the Universe         Universe is good      destroying Earth       is bad




      Hawking radiation                           Black holes will
       does not exist                            not destroy Earth

                                                  Black holes will
     Dr Azzeccagarbugli
                                                evaporate because
          says so
                                               of Hawking radiation

                             Dr Azzeccagarbugli       Now we are again justified
                          is not expert in physics
                                                      in believing that we should

                              He is a lawyer          run LHC   !
                                                                                   8
What s argumentation? (1)

[Prakken 2011] Argumentation is the process of supporting claims with
               grounds and defending them against attack.


[van Eemeren et al, 1996] Argumentation is a verbal and social activity
                         of reason aimed at increasing (or decreasing)
                         the acceptability of a controversial standpoint
                         for the listener or reader, by putting forward
                         a constellation of propositions intended
                          to justify (or refute) the standpoint
                          before a rational judge.


•  A framework for practical and uncertain reasoning able to cope
  with partial and inconsistent knowledge
     - philosophical roots: Aristotle, Toulmin (1958)
     - in AI: R.P. Loui (1987), J. Pollock (1987), G. Simari & Loui (1992)
                                                                             9
What s argumentation? (2)


The elements of an argumentation system


 •  The definition of argument
        (possibly including an underlying logical language +
         a notion of logical consequence)


 •  The notion of attack and defeat (successful attack) between arguments


 •  An argumentation semantics selecting acceptable (justified) arguments




                                                                        10
Definition of argument: several possibilities (1)

•  ASSUMPTION-BASED ARGUMENTATION

     Given a knowledge base (K, Ass)

                      Consistent theory   Set of assumptions

         ARGUMENT for p:

              (A, p) such that

                 - A " Ass
                 - A # K is consistent and entails p
                 - There is no A’$A such that A’ # K entails p

          ATTACKS to an argument: on its assumptions


                           [see Besnard&Hunter, Dung-Kowalski-Toni]
                                                                      11
Definition of argument: several possibilities (2)

•  ARGUMENT SCHEMES

     - correspond to recurring patterns of reasoning
     - have associated “critical questions”


Example: Expert Testimony                [WALTON 1996]

     E is expert on D
     E says P
     P is in D
        Therefore, P is the case

     Critical questions:
        Is E biased?
        Is P consistent with what other experts say?
        Is P consistent with known evidence?

                                                         12
Definition of argument: several possibilities (3)

•  ARGUMENT SCHEMES IN A MEDICAL APPLICATION

    Viability Scheme
       Organ O of donor D is available
       No contraindications are known for donating O to recipient R
       Therefore, organ O is viable

       CRITICAL QUESTIONS:
         Does donor D have a contraindication for donating organ O?

    Nonviability Scheme

       Donor D of organ O has condition C
       C is a contraindication are for donating O
       Therefore, organ O is nonviable


                                           [Tolchinsky et al, 2006]
                                                                      13
Definition of argument: several possibilities (4)

•  STABLE MARRIAGE PROBLEM
  - Arguments of the kind <Alice, John>
  - <Barbara, John> attacks <Alice, John> if John prefers Barbara to Alice

•  PLANNING
  - Plans as arguments (that a goal will be achieved)
  - Defeat between plans as attacks

………

 In general

    Arguments take different forms
                (domain-independent vs. domain dependent)


     Today examples will refer to rule-based approaches…

                                                                        14
Rule-based approaches

•  ARGUMENT

     a tree made up of rules of inference constructed from
     a set of premises to reach a conclusion


•  Two kinds of rules:                              A (0.7)
                                                            % ¬C (0.7)
           A % B:   deductive - indefeasible        B (0.9)
           A ! B:   non-deductive - defeasible
                                                     D (0.9) ! C (0.8)
•  A strength value may be associated to premises
  and rules, giving rise to argument strength


                           See [J.Pollock, 1992], [G. Vreeswijk, 1997], …

                                                                         15
Rule-based approaches (2)

Notion of conflict

 –  Rebutting:
      an argument attacks another one by denying its
      [possibly intermediate] conclusion

 –  Undercutting:
      an argument attacks the applicability of a
      defeasible rule of inference

 Notion of defeat
                                             A
      An argument ' defeats ( iff:              % ¬C
                                             B
       - ' undercuts (, or
       - ' rebuts ( and                        D!C
         ' is not weaker than (
                                            E!(D&C)
                                                       [Pollock 92]
                                                                      16
Rule-based approaches (3)

EXAMPLE


             REBUTTING DEFEAT

                                     UNDERCUTTING
              It’s       It’s not       DEFEAT       Bob
            raining      raining                is unreliable




          Smith says     Bob says                 Bob is drunk
          it’s raining    it’s not
                          raining
                                                                 17
The ASPIC framework

•  One result of the European ASPIC Project (2004-2006)
•  Generalizes Pollock’s rule-based approach in several respects:
  - any logical language (and an associated ‘contrariness function’
    generalizing classical negation) can be adopted
  - can be instantiated by a partial preorder on defeasible rules
  - premises are distinguished into necessary, ordinary and assumption
    premises (ordinary and assumption premises partially preordered)
  - a partial preorder is assumed between arguments
•  Besnard & Hunter’s approach, Pollock’s system… can be obtained as
 instances of ASPIC framework
•  See [H. Prakken, “An abstract framework for argumentation with
  structured arguments”, Argument and Computation, 2010] for details.


                                                                       18
Argumentation in the context of MAS (1)

Advantageous features

   •  Several kinds of arguments can be represented
     - epistemic reasoning
     - practical reasoning
   •  Able to handle uncertain and partial knowledge
     - nonmonotonic notion of warrant:
          1) wrt further information
          2) wrt further reasoning steps (anytime reasoning framework)
   •  A natural representation + justification of choices
    (in terms of argument , rebuttal , counterargument …)
   •  Argumentation has a dialogical side
    (in terms of argument , attack , defence …)


                                                                     19
Argumentation in the context of MAS (2)

The uses of argumentation (examples)


                  AUTONOMOUS              MULTI-AGENT
                  REASONING               INTERACTION
 EPISTEMIC - Belief Revision
 REASONING (arguing over beliefs)
                  - Trust management
                    (arguing over other   - Dialectics in
                    agents reputation)      Multiagent Interaction
 PRATICAL  - Decision making
 REASONING (arguing about the
                   expected value of
                   possible actions)
                                                                     20
What s argumentation? (3)


The elements of an argumentation system


 •  The definition of argument
        (possibly including an underlying logical language +
         a notion of logical consequence)


 •  The notion of attack and defeat (successful attack) between arguments


 •  An argumentation semantics selecting acceptable (justified) arguments




                                                                        21
What s abstract argumentation?

Usually “abstract” stands for a difficult thing… Here it means “simple”!

The elements of an argumentation system


 •  The definition of argument
        (possibly including an underlying logical language +
         a notion of logical consequence)


 •  The notion of attack and defeat (successful attack) between arguments


 •  An argumentation semantics selecting acceptable (justified) arguments




     Abstract argumentation focuses on this aspect

                                                                           22
Dung s argumentation framework

                                                         [Dung ’95]
  AF = <A, %>
                           attack (or defeat) relation
                           [unspecified definition]

                            Arguments [origin and structure not specified]


•  Graphical representation as a directed graph [defeat graph], e.g.

               Representation of LHC example




               Representation of weather example




                                                                       23
Dung s argumentation framework (2)

 So, what remains to be done?

  ARGUMENT EVALUATION:

   GIVEN AN ARGUMENTATION FRAMEWORK,
   DETERMINE THE JUSTIFICATION STATE
   (ALSO CALLED DEFEAT STATUS) OF ARGUMENTS,
   IN PARTICULAR: WHAT ARGUMENTS EMERGE UNDEFEATED
          FROM THE CONFLICT, I.E. ARE ACCEPTABLE?




                                                     24
Argumentation semantics

•  Specification of a method for argument evaluation, or of
  criteria to determine, given a set of arguments, their defeat status




                                 Semantics
Argumentation Framework                          Defeat status


                                               Undefeated

                 Defeat status                 Defeated

                                               Provisionally Defeated

                                                                         25
Labelling vs. extension-based semantics

LABELLING-BASED SEMANTICS

   - Based on the notion of labelling
     [assignment to each argument of a label from a predefined set]
   - Specifies how to derive from an argumentation framework
     a set of labellings
   - Justification of arguments derived from the set of labellings


EXTENSION-BASED SEMANTICS

   - Less general (at least in theory), but more common
    kind of semantics
   -  Based on the notion of extension
      [set of arguments collectively acceptable ]



                                                                      26
Extension-based semantics

                         Semantics S



Argumentation framework AF             Set of extensions   S(AF)




                                                                   27
From extensions to defeat status

Set of extensions    S(AF)                 Defeat/Justification Status




A common definition

   •  Skeptically justified argument: belongs to all of the extensions
   •  Credulously justified argument: belongs to at least one
   •  Indefensible argument: does not belong to any extension

                                                                         28
Unique-status vs. multiple-status semantics


             Unique-Status Semantics

                                          Unique extension: empty set
                      ')        ()
                                          ' and ( directly unjustified
                                                     (provisionally defeated)


')      ()
             Multiple-Status Semantics


               ')          ()        ')        ()


                    ! ' and ( unjustified (provisionally defeated)




                                                                                29
Relationship between labelling and
          extension-based approaches
•  Almost all approaches adopt the set {IN, OUT, UNDEC}
  - IN = belonging to the extension
  - OUT = attacked by the extension
  - UNDEC= not belonging to nor attacked by the extension

                        Unique-Status Semantics


                                ')          ()

                           UNDEC           UNDEC
          ')      ()
                         Multiple-Status Semantics


                           ')         ()         ')    ()

                          IN         OUT         OUT   IN

                                                            30
The core of Dung’s theory: complete “semantics”

Acceptability
 ' acceptable w.r.t. (“defended by”) S                 '
    •  all attackers of   ' are attacked by S


Admissible set S
                                                        S
    •  conflict-free
    •  every element acceptable w.r.t. S
      (defends all of its elements)                Complete semantics


                                                            Complete
                                   IF
                                                            extension
                             also includes all
                             acceptable elements
                             w.r.t. itself         All traditional semantics
                                                   select complete extensions
                                                                           31
Complete “semantics”: examples

Chain
                               Admissible sets:
                                    ø,
                                     {'}, {', *}
    ')          ()        *)   Only one complete extension:

                                     CO(AF) = {{', *}}


Nixon Diamond


                     ')        ()        All admissible sets
                                         are complete


    ')          ()                        CO(AF) =
                     ')        ()
                                            {   ø, {'}, {(} }

                     ')        ()
                                                                32
Complete “semantics”: examples (2)

Nixon Diamond + node

                               Admissible sets:
    ')        ()       *)        ø, {'}, {(}, {', *}


                                          CO(AF) = {

               ')       ()       *)          ø


     CO(AF)    ')       ()       *)        {', *},



               ')       ()       *)        {(} }



                                                       33
The Grounded Semantics: a unique status approach


                            Grounded extension GE(AF):

                                   Least complete extension

                                 included in all extensions
                                 of any traditional semantics

                                 Grounded semantics is
                                 the “most skeptical” one



                                      Undefeated

            Defeat status             Defeated

                                      Provisionally Defeated
                                                            34
Grounded semantics: examples

Chain


    ')          ()       *)         GE(AF) = {', *}


Nixon Diamond



    ')          ()     GE(AF) = ø


Nixon Diamond + node



    ')          ()       *)     GE(AF) = ø

                                                      35
Floating arguments: a problem for grounded semantics

•  Actually, grounded semantics is polynomially computable
•  But sometimes a more discriminative behavior is desirable

THE CASE OF FLOATING ARGUMENTS


     ()                                         ()

                *)         +)      VS                        *)    +)

     ')                                         ')

      Grounded Semantics                             What we (may) want


 •  A problem for all possible unique status approaches

               Let us consider multiple status approaches!


                                                                        36
Stable Semantics

Stable extension = conflict-free set attacking all outside arguments

THE CASE OF FLOATING ARGUMENTS

         ()                                ()

                    *)         +)                       *)         +)

         ')                                ')


       ST(AF) = { {',    +}, {(, +} }    !      + is justified

ODD-LENGTH CYCLES: A PROBLEM FOR STABLE SEMANTICS
         ')
                          No stable extension exists!
                    ()

          *)              (and also imposing    ø is not satisfactory)
                                                                         37
Stable Semantics: an unsatisfactory patch


Stable extensions =
        - conflict-free sets attacking all outside arguments, if there is one
        - {ø}, otherwise




                                   '1)


                                                   '3)



        ()            *)           '2)



         ST(AF) = {ø }      !     ( NOT justified!!!

                                                                           38
Preferred semantics

Stable extensions are maximal complete extensions
        •  conflict-free: by definition
        •  admissible: every argument attacking an extension is outside
                         ! attacked by the extension itself
        •  maximal: no argument can be included!


Preferred semantics         [P.M. Dung, 95]

Preferred extension

  Maximal complete extension = max Set:
     •  is conflict-free
     •  defends all of its elements




                                                                          39
Preferred semantics and floating arguments


                           ()
                                      *)        +)
()                                                          ()
        *)        +)       ')
                                                                     *)    +)
')                          ()
                                                            ')
                                      *)        +)
                           ')


PR(AF) =       ST(AF) = { {',    +}, {(, +} }        !    + is justified


                                           ()
             Grounded semantics:                     *)    +)
                                           ')

                                                                                40
Preferred semantics and odd-length cycles


')
                      PR(AF) = {ø}
           ()                               A big difference, isn’t it?
                       ST(AF) =   ø
*)

                      GE(AF) = {ø}


No argument justified w.r.t. grounded and preferred semantics

          •  As stable semantics, preferred semantics handles
                the case of floating arguments
                (differently wrt grounded semantics)
          •  W.r.t. stable semantics it behaves “better”
                in the case of odd-length cycles
                (as the grounded semantics)

           So, what remains to be done?
                                                                          41
Semi-stable semantics (1)

•  Stable semantics
    - clashes in some cases (odd-length cycles), however:
    - a widely applied approach (default logic, stable models of
      logic programming, answer set programming, etc.)
    - a very credulous approach:
         stable extensions are preferred but not viceversa
         ! justified arguments w.r.t. stable semantics are a
            (sometimes strict) superset of arguments justified
           w.r.t. preferred semantics, e.g.

      ,)                                 ,)

               *)      ')      ()                *)         ')      ()

      +)                                 +)

                        PR(AF)={{',   !}, {"}}    ST(AF)   = {{', !}}
                                                                         42
Semi-stable semantics (2)

•  Aims at guaranteeing existence of extensions             [Verheij’96,
            (differently from stable semantics)              Caminada’06]
 + coinciding with stable semantics when stable extensions exist
            (differently from preferred semantics)

•  Definition:
    E-    SST(AF)   iff
    E is a complete extension such that (E U {'| E% '}) is maximal

 •  Main properties:
     -  A semistable extension always exists (in the finite case!)
        since a maximization requirement replaces “aggressive attack”
     -  If a stable extension E exists, then (E U {'| E% '}) includes
        all arguments, therefore semistable extensions # stable extensions
     -  In any case, semistable extensions are preferred extensions, but
        the opposite is not always true
                                                                         43
Semi-stable semantics: examples

Example for existence
        ')
                        The unique admissible set is empty
                ()        ! trivially maximizes (E U {'| E% '} )

         *)

Example for backward compatibility
       (and difference w.r.t. preferred semantics)


         ,)
                                        PR(AF)={{',   !}, {"}}
                 *)      ')      ()
                                         SST(AF)={{',   !}} )=   ST(AF)
         +)




                                                                     44
CF2 semantics: motivation


Preferred/stable/semistable semantics and cycles


                                              ')
    ')        ()
                                                         ()

    ')        ()                               *)




                                               ')
    ')        ()
                                                         ()

                                               *)


         A different treatment for even and odd-length cycles.
         Is it just a matter of symmetry and elegance?
                                                                 45
Preferred/Semistable Semantics and cycles

                               PR(AF) =
')      ()        +1)    +2)
                                {{', +1}, {', +2},

             VS                           {(, +2} }




                                                      46
Preferred/Semistable Semantics and cycles

                               PR(AF) =
')      ()         +1)   +2)
                                {{', +1}, {', +2},

             VS                           {(, +2} }

()
        *)         +1)   +2)   PR(AF) = {{+2}}

')
              VS




                                                      47
Preferred/Semistable Semantics and cycles

                                                 PR(AF) =
       ')        ()          +1)        +2)
                                                 {{', +1}, {', +2},

                        VS                                 {(, +2} }

       ()
                 *)          +1)       +2)       PR(AF) = {{+2}}

       ')
                        VS

            ()
                                                 PR(AF) =
                                               {{', *, +2},
  ')             *)          +1)       +2)
                                                {(, +, +1}, {(,   +, +2} }
            +)
NOTE: grounded semantics yields the empty set in all cases             48
Pollock example revisited (1)




   Jones                       Smith           It’s       It’s not
 unreliable                  unreliable      raining      raining


                  Rob
               unreliable




  Rob says     Smith says    Jones says    Smith says     Bob says
Jones unrel.   Rob unrel.   Smith unrel.   it’s raining    it’s not
                                                           raining



                                                                  49
Pollock example revisited (2)

               Fred says
              Jones unrel.




                 Jones
               unreliable

  Fred                          Smith           It’s       It’s not
unreliable                    unreliable      raining      raining


                  Rob
               unreliable




 Rob says     Smith says      Jones says    Smith says     Bob says
Fred unrel.   Rob unrel.     Smith unrel.   it’s raining    it’s not
                                                            raining
                                                                   50
Preferred Semantics and Floating Arguments again…

      ()

                                                [ two preferred
                     *)           +)                extensions]


      ')

                     VS

      ()
                                             [empty set is the unique
                *)          +)         .)       preferred extension]

     ')


NB: same behavior for semistable semantics, stable semantics clashes,
    grounded semantics yields the empty set in both cases
                                                                        51
Strongly connected components (SCCs)

     Equivalence classes under the relation of
     path-equivalence (mutual reachability)


()
         *)        .1)     .2)
')
                                   ()
                                         *)      .1)   .2)
                                   ')
 ()
              *)     .1)     .2)
 ')
                                                             52
Strongly connected components (SCCs)

          SCCs form an acyclic graph

                   S3                           S6

S1
                  S4




S2
                                S5
                                                        S7
                        S1 and S2 are initial SCCs
                        S1 is sccparent of S3, S4 and
                        S5
                        all other SCCs precede S7
                                                             53
CF2 semantics: the definition


E-   CF2(AF)   iff:

     - E - MCF(AF)                         if |SCCSAF| = 1
     - / S - SCCSAF
        (E0S) - CF2(AF       UP_AF(S,E))   otherwise




                        S

                                           UP_AF(S,E)


                                                             54
CF2 semantics and odd-length cycles (1)


                 ()

                             *)

                 ')


    ()                ()                ()
            *)                    *)         *)
    ')                ')                ')
           Maximal conflict-free sets

                                                  55
CF2 semantics and odd-length cycles (2)



      ()
                 *)     .1)     .2)
      ')
{*,.2}, {',.1}, {',.2}, {(,.1}, {(,.2}
                                            Yields several extensions
                                            ! all arguments not justified
                                               in both cases
       ()
 ')         *)        .1)     .2)
       +)    {',*,.2}, {(,+,.1}, {(,+,.2}



                                                                        56
Floating arguments with a three-length cycle


()
         *)       +)       .)
')
                                       ()
()
                                              *)        +)   .)
         *)       +)       .)
                                       ')
')

()
         *)      +)        .)
')
     Extensions: {*,.}, {',.}, {(,.}    Defeat status
                                                              57
since we do not want problems in one in relation toknowledge base to affect other,
                             advantageous part of the consistency requirements, as explained in the following.
  ly unrelated parts of the knowledge base. generates an argumentation framework based on a set of propositional formul
                                Suppose one Stable semantics is therefore not an option.
   semantics have to be admissibility based? That is,rules desirable that each the propositional formulas express informatio
                             P and a set of defeasible is it D. The idea is that extension
  an admissible (or even complete) one? Again, it ofthe defeasible rules an ultimate of thumb that can be subject
                             that isA problem is difficult to provide express(1)
                                     beyond doubt and CF2 semantics rules
neral: one has to refer to specific contexts. In particular, following knowledge base:
                             exceptions. Now consider the in the context of instantiated
  nerated from an underlying logical knowledge base, admissibility can be regarded as
                                                              P ¼ fjw; mw; sw; :ðjt ^ mt ^ stÞg
 s in relation to consistency requirements, as explained in the following.
 ne generates•  Considering some examples with structuredmw ) mt; sw ) it turns out that
                an argumentation framework based on a set ¼ fjw ) jt; arguments, stg
                                                              D of propositional formulas
  f defeasible rules D. The idea is that the propositional formulas express information and Suzy want to go cycling on
                  conflict-freenessexpress not entail consistency, e.g. Mary,
                                This example can be interpreted as follows: John,
nd doubt and the defeasible rules doesthat John wants to thaton thebe subject tois a reason to believe that John will b
                             tandem. The fact rules of thumb get can tandem (jw)
                                                                    An introduction to argumentation semantics
Now consider the following knowledge base: The same holds for Mary and Suzy. However, since the tandem only has tw      407
                             on the tandem (jt).
                         seats, they :ðjt ^ mt ^ it with the three of them: :(jt 4 mt 4 st). From this knowledge base, w
                   P ¼ fjw; mw; sw;cannot be onstÞg                              A5
                         can then construct the 10 following arguments, based on an argument construction scheme
                   D ¼ fjw ) jt; mw ) mt; sw ) stg Amgoud (2007) and Prakken (2010):
                         presented in Caminada and
                              A1 5 :(jt 4 mt 4 st)
 ple can be interpreted as follows: John, Mary, and Suzy want to go cycling on a
                                                                  A2             A8                    A3
                              A 5 tandem (jw) is a reason to believe that John will be
fact that John wants to get on2thejw
                              A3 5 mw
m (jt). The same holds for Mary and Suzy. However, since the tandem only has two
 nnot be on it with the three A4 5 sw :(jt 4 mt 4 st). From this knowledge base, we
                              of them:
                                                                              A1
                              A5 5 A2 ) jt on an argument construction scheme as
 struct the 10 following arguments, based
                                                                    A10                           A9
Caminada and Amgoud (2007)5 A3 ) mt
                              A6 and Prakken (2010):
                              A7 5 A4 ) st
4 mt 4 st)                    A8 5 A6, A7, A1 - :jt
                              A9 5 A5, A7, A1 - :mt                           A4
                                                                 A7                                         A6
                              A10 5 A5, A6, A1 - :st

) jt                          Assuming the principle of restrictednot enough to obtain consistent conclusions
                                         Figure 20 Conflict-freeness is
                                                                        rebutting23 it would then follow that A8 attacks A5, A9, an
                           A10, that A9 attacks A6, A8, and A10, and that A10 attacks A7, A8, and A9. This yields th     58
 mt
                           argumentation also semi-stable and preferred extensions). It should be mentioned that the sets of conclusions
                                          framework of Figure 20.
A problem of CF2 semantics (2)


•  By slightly complicating example, one can find a CF2 extension
  which is not consistent
•  On the other hand, admissibility entails consistency
            (proved by Caminada & Amgoud 2007, AIJ)




        ADMISSIBILITY CAN BE A DESIRED REQUIREMENT




                                                                    59
TO CONCLUDE…


                      GROUNDED

            PRUDENT              STABLE
 STAGE                                    PREFERRED

             ROBUST          SEMISTABLE

TOLERANT                                      CF2

           SUSTAINABLE           IDEAL




EACH SEMANTICS HAS ITS OWN ROLE…
  … WHICH ONE IS A GOOD RESEARCH QUESTION…!
                                                      60
What about general principles?

Here we consider only some semantics - see [Baroni & Giacomin ’06]

                          Grounded Preferred           CF2   Semistable
   CF-principle               Yes          Yes         Yes      Yes
   Admissibility              Yes          Yes         No       Yes
   Reinstatement              Yes          Yes         No       Yes
   Weak reinstatement         Yes          Yes         Yes      Yes
   CF-reinstatement           Yes          Yes         Yes      Yes
   I-maximality               Yes          Yes         Yes      Yes
   Directionality             Yes          Yes         Yes      No
   Weak Skepticism            Yes           No         Yes      No
   Adequacy                [all forms]

   Weak Resolution            No           Yes         No       Yes
   Adequacy                              [all forms]



                                                                          61
Applications and principles


                 PRINCIPLES




               TO BE      TO BE
              STUDIED   DEEPENED



APPLICATION
  DOMAINS                          SEMANTICS

                                               62
Semantics and attitude
SKEPTICAL REASONING                              CREDULOUS REASONING
      E                                                E
E1     W E2 :                                    E1    C E2 :
     / E2 -E2,   1 E1 -E1 : E1" E2                    / E1 -E1,   1 E2 -E2 : E1" E2




                                                                                  63

             Fig. 3.      S
                              +,
                                   S
                                       →   and    S   relations for any argumentatio
MANY THANKS
FOR YOUR KIND ATTENTION




                          64
Selected references (1)
Landmark argumentation papers and books

 S. Toulmin, “The Uses of Argument”
          Cambridge University Press, 1958.

 R. P. Loui, “Defeat Among Arguments: a System of Defeasible Inference”,
 Computational Intelligence, vol. 3(3), 1987.

 J. Pollock, “Defeasible Reasoning”,
 Cognitive Science, vol. 11(4), 1987.

 G. Simari & R. P. Loui, “A mathematical treatment of defeasible reasoning and
 its implementation , Artificial Intelligence, vol. 53(2-3), 1992.

Argumentation surveys

 H. Prakken & G.A.W. Vreeswijk, “Logics for Defeasible Argumentation”,
 in Handbook of Philosophical Logic, 2nd Edition, Kluwer Academic Publishers, 2001.

 C.I. Chesnevar, A.G. Maguitman, R.P. Loui, “Logical models of argument”,
 ACM Computing Surveys, vol. 32(4), 2000.
                                                                                 65
Selected references (2)
Argumentation semantics survey
  P. Baroni, M. Caminada, M. Giacomin “An introduction to argumentation
  semantics , The Knowledge Engineering Review, vol. 26(4),2011.


Books

 D. Walton, “Fundamentals of critical argumentation ,
              Cambridge University Press, 2006.

 P. Besnard & A. Hunter, “Elements of Argumentation , MIT Press, 2008.

 “Argumentation in Artificial Intelligence , edited by I. Rahwan and G. R. Simari,
 Springer, 2009.


Dung s influential paper on abstract argumentation

 P.M. Dung, “On the Acceptability of Arguments and Its Fundamental
 Role in Nonmonotonic Reasoning, Logic Programming, and n-Person Games ,
 Artificial Intelligence, vol. 77(2), 1995.
                                                                                 66
Selected references (3)
Semantics
 P. Baroni, M. Giacomin, G. Guida, “SCC-recursiveness: a general schema
 for argumentation semantics”
 Artificial Intelligence, vol. 168(1-2), 2005.
 B. Verheij, “Two approaches to dialectical argumentation:admissible sets
 and argumentation stages”, Proc. of the 8th Dutch Conference on
 Artificial Intelligence, 1996

 M. Caminada, “Semi-Stable Semantics”, Proc. of 1st International Conference on
 Computational Models of Arguments (COMMA 2006), 2006

 P.M. Dung, P. Mancarella, F. Toni, “A dialectic procedure for sceptical,
 assumption-based argumentation”, Proc. of 1st International Conference on
 Computational Models of Arguments (COMMA 2006), 2006

 S. Coste-Marquis, C. Devred, P. Marquis, "Prudent Semantics for Argumentation
 Frameworks", Proc. of 17th IEEE International Conference on Tools with
 Artificial Intelligence (ICTAI 2005), 2005

                                                                              67
Selected references (4)

Semantics

  H. Jakobovits & D. Vermeir, "Robust Semantics for Argumentation Frameworks",
  Journal of Logic and Computation 9(2), 1999

 P. Baroni, M. Giacomin, G. Guida, “SCC-recursiveness: a general schema
 for argumentation semantics”
 Artificial Intelligence, vol. 168(1-2), 2005.

 P. Baroni, M. Giacomin, “Resolution-based argumentation semantics”,
 Proc. of 2nd International Conference on Computational Models of Arguments
 (COMMA 2008), 2008

 G.A. Bodanza, F.A. Tohmé, “Two approaches to the problems of self-attacking
 arguments and general odd-length cycles of attack”
 Journal of Applied Logic, to appear.

  P. Baroni, P. Dunne, M. Giacomin, “Computational Properties of Resolution-based
  Grounded Semantics”, IJCAI 2009, to appear.

                                                                               68
Selected references (5)

General criteria for semantics evaluation and comparison

  M. Caminada & L. Amgoud, “On the evaluation of argumentation formalisms”,
  Artificial Intelligence, vol. 171(5-6), 2007.

  P. Baroni, M. Giacomin, G. Guida, “On principle-based evaluation of
  extension-based argumentation semantics”, Artificial Intelligence,
  vol. 171(10-15), 2007.

  P. Baroni, M. Giacomin, Skepticism relations for comparing argumentation
  semantics , International Journal of Approximate Reasoning, vol. 50(6), 2009.




                                                                                  69
!"#$%&'()*+',-+",!#&'(,.+/0&*&1,
           2)"(,33,

4%33(5()(%$/&6(%7/5($&         !"#"$%&'())%*%&
 8$(9"#3(*%:&;(&<#"37(%&   +,-+.&!/01(%&.$20/)(3&
!"#$%&'()*+',-+",:&/010+',;)<0'#,



!"#$%&'()*+',-+",4&50&-1,6&7010+',




                                           !"#$%&'()*+',-+",6&)1+'0'#,,
                                                          )8+$(,9"$1(,
:&/010+',;)<0'#,
!"#$%&'()*+',-+",=&/010+',%)<0'#,
•  :&/010+',%)<0'#>,
  –  .&5&/(,(?&,8&1(,+",#++=,)/*+'1@,
  –  !/*+'1,-&)1085&,)%+'#,)5(&"')*7&1A,


•  B&,'&&=,(+,<'+C,
  –  3'-+"%)*+',D0-,)7)05)85&E,)8+$(,(?&,/$""&'(,1()(&,
     +-,(?&,C+"5=@,
  –  F+'1&G$&'/&1,+-,H+(&'*)5,)/*+'1A,,
!"#$%&'()*+',-+",=&/010+',%)<0'#,
•  2"+85&%1>,
  –  !7)05)85&,0'-+"%)*+',0'/+%H5&(&,
  –  !7)05)85&,0'-+"%)*+',H&"7)=&=,C0(?,$'/&"()0'(I,



  =/>&5?71&*1"&7/$3"@?"$7"3&/A&%$&%72/$&B*&*1"&
   0#"A"#"$7"3&/A&*1"&;"7(3(/$&5%C"#D&
!"#$%&'()*+',-+",=&/010+',%)<0'#,
•  F5)110/)5,=&/010+',(?&+"I,
  –  2"0'/0H5&1,-+",/+%H)"0'#,=0J&"&'(,)5(&"')*7&1A,
  –  3'H$(1>,
     •  .&(,+-,/)'=0=)(&,)/*+'1@,
     •  K$'/*+',(+,)11&11,(?&,7)5$&,+-,(?&0",/+'1&G$&'/&1@,
     •  F+%H5&(&,+",H)"*)5,0'-+"%)*+',)8+$(,1()(&,+-,C+"5=A,
  –  L$(H$(>,H"&-&"&'/&,"&5)*+',8&(C&&',)/*+'1A,
!"#$%&'()*+',-+",=&/010+',%)<0'#,
•  M'+C5&=#&,)'=,2"&-&"&'/&1>,
  –  :01("08$*+',-$'/*+',)11&110'#,H5)$108050(I,+-,(?&,
     1()(&1,+-,(?&,C+"5=@,
  –  N*50(I,-$'/*+',&'/+=0'#,H"&-&"&'/&1,(+,&1*%)(&,
     (?&,#++='&11,+-,),/+'1&G$&'/&A,
•  2"0'/0H5&1,$1&=,(+,/+%H)"&,H)0"1,+-,
   )5(&"')*7&1A,
!"#$%&'()*+',-+",=&/010+',%)<0'#,
•  !')5I*/)5,&OH"&110+'1,1$%%)"0P&,(?&,
   =&/010+'A,

•  Q)"=,(+,$'=&"1()'=,C?I,),H"+H+1&=,
   )5(&"')*7&,01,#++=@,+",8&R&"A,

•  B&,'&&=,)',)HH"+)/?,(+,$'=&"1()'=,(?&,
   $'=&"H0''0'#,+-,(?&,&7)5$)*+'A,
!"#$%&'()*+',-+",=&/010+',%)<0'#,
•  B?I,$10'#,)"#$%&'()*+'S,
  –  TOH5)')(+"I,H+C&"@,
  –  :&/010+'1,8)1&=,+',)"#$%&'(1,)'=,,,,,,,,,,,,,,
     /+$'(&"U)"#$%&'(1@,
  –  !85&,(+,&OH5)0',!"#,),/?+0/&,?)1,8&&',%)=&A,
9C+,)5(&"')*7&,)HH"+)/?&1,
•  !%#+$=@,2")=&A,83($E&%#E?5"$*3&A/#&5%C($E&
   %$;&"F0)%($($E&;"7(3(/$3G,!"*-A,3'(&55A,VWXDXUYE>,
   YVXUYXZ,D[]EA,
•  M)<)1@,;+")0*1A,.#E?5"$*%2/$&H%3";&;"7(3(/$&
   5%C($E&A/#&%?*/$/5/?3&%E"$*3G&3',2"+/1,+-,
   !!;!.,[X>,^^XU^]A,
!%#+$=,_,2")=&,)HH"+)/?,
•  :&`'&,),D/+%H5&(&E,H"&U+"=&"0'#,+',),1&(,+-,
   H+11085&,+H*+'1,:A,
•  !"#$%&'()*+'U8)1&=,=&/010+',%)<0'#>,
  –  F+'1("$/*'#,)"#$%&'(1,0',-)7+"a)#)0'1(,1()(&%&'(1,
     DH&"()0'0'#,(+,8&50&-1,+",=&/010+'1E@,
  –  T7)5$)*'#,(?&,1("&'#(?,+-,&)/?,)"#$%&'(@,
  –  :&(&"%0'0'#,(?&,=0J&"&'(,/+'b0/(1,)%+'#,)"#$%&'(1@,
  –  T7)5$)*'#,(?&,)//&H()8050(I,+-,)"#$%&'(1@,,
  –  F+%H)"0'#,=&/010+'1,+',(?&,8)101,+-,"&5&7)'(,
     c)//&H(&=d,)"#$%&'(1A,
$%&'()*+
•  ,*-./.012,?)70'#,),1$"#&"I,D1#E,+",'+(,De1#EA,
•  310!)*45*2,(?&,H)*&'(,?)1,/+5+'0/,H+5IH1A,
•  9?&,M4,/+'()0'1,(?&,-+55+C0'#,0'-+"%)*+'>,
    –  ?)70'#,),1$"#&"I,?)1,10=&U&J&/(1f,
    –  '+(,?)70'#,1$"#&"I,)7+0=1,?)70'#,10=&U&J&/(1f,,
    –  C?&',?)70'#,),/)'/&"@,?)70'#,),1$"#&"I,)7+0=1,5+11,+-,50-&f,
    –  0-,),H)*&'(,?)1,/)'/&",)'=,?)1,'+,1$"#&"I@,(?&,H)*&'(,C+$5=,5+1&,?01,
       50-&f,
    –  (?&,H)*&'(,?)1,/+5+'0/,H+5IH1f,
    –  ?)70'#,/+5+'0/,H+5IH1,%)I,5&)=,(+,/)'/&"f,

    g+)51,50<&>,c'+,10=&,&J&/(1d,)'=,c(+,'+(,5+1&,?01,50-&dA,,

    ;+"&,0%H+"()'(,-+",?0%,(+,'+(,5+1&,?01,50-&,(?)',(+,'+(,?)7&,10=&,&J&/(1A,
9?&,-")%&C+"<,
•  !81(")/(,)"#$%&'(U8)1&=,-")%&C+"<A,
•  9C+,1(&H1>,
  –  .(&H,V,
     •  !"#$%&'(1,-+",8&50&-1@,
     •  !"#$%&'(1,-+",+H*+'1@,
     •  4$05=0'#,)'=,&7)5$)*+',$'=&",/5)110/)5,1&%)'*/1A,
  –  .(&H,[,
     •  F+%H)"01+',+-,H)0"1,+-,+H*+'1,$10'#,=&/010+',
        H"0'/0H5&1A,
9?&,-")%&C+"<,
•  9?"&&,/5)11&1,+-,=&/010+',H"0'/0H5&1>,
  –  N'0H+5)">,,
     •  L'5I,)"#$%&'(1,H"+1@,
     •  L'5I,)"#$%&'(1,/+'1A,
  –  40H+5)",
     •  4+(?,(IH&1A,
  –  h+'UH+5)",
     •  !##"&#)*+',+-,H"+1,)'=,/+'1,)"#$%&'(1,0'(+,%&()U
        )"#$%&'(A,
9?&,-")%&C+"<,
•  :&/010+'1,1$HH+"(&=,8I,)"#$%&'(1>,
  –  &%H?)10P0'#,0(1,H+10*7&,/+'1&G$&'/&1@,,
  –  #+)51,1)*1`&=@,
  –  "&i&/*+'1,)7+0=&=A,
•  :&/010+'1,)R)/<&=,8I,)"#$%&'(1>,,
  –  &%H?)10P0'#,0(1,'&#)*7&,/+'1&G$&'/&1@,,
  –  %011&=,#+)51@,
  –  /&"()0',"&i&/*+'1A,
9C+,)5(&"')*7&,)HH"+)/?&1,
•  !%#+$=@,2")=&A,83($E&%#E?5"$*3&A/#&5%C($E&
   %$;&"F0)%($($E&;"7(3(/$3G&!"*-A,3'(&55A,VWXDXUYE>,
   YVXUYXZ,D[]EA,
•  M)<)1@,;+")0*1A,.#E?5"$*%2/$&H%3";&;"7(3(/$&
   5%C($E&A/#&%?*/$/5/?3&%E"$*3G&3',2"+/1,+-,
   !!;!.,[X>,^^XU^]A,
M)<)1,_,;+")0*1,)HH"+)/?,
•  ;+=$5)",)#&'(,)"/?0(&/($"&>,
  –  ;+=$5&1,=&=0/)(&=,(+,)#&'(j1,/)H)8050*&1@,
  –  ;+=$5&1,)"&,0'=&H&'=&'(A,
•  L7&")55,8&?)70+$",+-,(?&,)#&'(>,
  –  3'(&")/*+'1,)%+'#,=0J&"&'(,%+=$5&1@,
  –  3'(&")/*+'1,)%+'#,(?&0",1&H)")(&,=&/010+'1@,
  –  :&508&")*+',H"+/&11,C0(?0',&)/?,%+=$5&A,
9?&,%+=&5,
•  !"#$%&'()*7&,=&508&")*+',%+=$5&A,
•  k)"0+$1,=&/010+',%)<0'#,("&)(&=,$'0-+"%5IA,
•  Q0#?,5&7&5,+-,)=)H()8050(I,C?&',(?&,
   &'70"+'%&'(,/?)'#&1A,
•  T'/+%H)110'#,(?&,0'b$&'/&,+-,
  –  ,=0J&"&'(,"&5)*7&,"+5&1,+-,0'(&")/*'#,)#&'(1@,)'=,
  –  ,(?&,/+'(&O(,+-,(?&,H)"*/$5)",0'(&")/*+'A,
9?&,%+=&5,
•  6+5&1,)'=,/+'(&O(,)1,=I')%0/,H"&-&"&'/&1,+',
   (?&,=&/010+',H+50/0&1A,
•  6+5&1,)'=,/+'(&O(,"&H"&1&'(&=,0',(C+,%+=$5)",
   H)"(1A,
•  !0%0'#,(+,H"+70=&,(?&,)#&'(,C0(?,"+8$1('&11>,
  –  0'/+%H5&(&,0'-+"%)*+',-"+%,&'70"+'%&'(,!,
     )8=$/*+',C0(?0',)"#$%&'()*+',-")%&C+"<f,
  –  )#&'(,)85&,(+,=&508&")(&,+',)5(&"')*7&,/?+0/&f,
  –  )#&'(,()<&,=&/010+'1,/+'=0*+')5,+',)11$%H*+'1,
     )8+$(,&'70"+'%&'(f,
9?&,%+=&5,
•  2&"1+')50(I,%+=$5&,
  –  8)1&=,+',(?&,1)%&,)"#$%&'()*+',-")%&C+"<,
     $1&=,-+",(?&,+(?&",%+=$5&1@,
  –  "&H"&1&'*'#,(?&,H&"1+')50(I,+-,)',)#&'(,)1,),
     =&/010+',H+50/I,)//+"=0'#,(+,'&&=1,)'=,
     %+*7)*+'1,+-,)#&'(1@,
  –  )==0*+')5,i$=#%&'(,+',=&/010+',H"+85&%1,+-,)'I,
     +'&,+-,(?&,+(?&",%+=$5&1,=&H&'=0'#,+',7)"0+$1,
     '&&=1A,
9?&,%+=&5,
•  9?"&&,5&7&51,0',)',)#&'(j1,(?&+"I>,
  –  K0"1(,5&7&5>,"$5&1,9,(?)(,"&-&",=0"&/(5I,(+,(?&,1$8i&/(,
     =+%)0',+-,(?&,)#&'(,D678*-9:)*;*)+,*-./.01+<=)*/+
     0>+9"*+&5*19?+
  –  .&/+'=,)'=,(?0"=,5&7&5>,"$5&1,"&5)(&,(+,(?&,H+50/I,
     $'=&",C?0/?,(?&,)#&'(,$1&1,?01,+8i&/(U5&7&5,
     =&/010+',"$5&1,)//+"=0'#,(+,"+5&1,@AB+C+<0)*+
     AB.0B.D*/?E+)'=,/+'(&O(,@A-+:+F019*%9+AB.0B.D*/?G,
$%&'()*+
"V,>,"&G$&1(l?&5HD!@9@!VE,m,'&&=l/++H&")*+'D!@9E@,
   "&5&7)'(D9@!VE,
"[,>,e"&G$&1(l?&5HD!@9@,!VE,m,"&G$&1(l?&5HD!@9@,![E@,,
  ,!V,n,![,,
6V,>,?lHD"VD!@9@!VE@"VD!@9@![EE,m,%)')#&%&'(l()1<
   D9E@%)')#&"D!VE,,
6[,>,?lHD"VD!@9@,![EE@,"VD!@9@,!VEE,m,(&/?'0/)5l()1<
   D9E@&OH&"(D![E,,
FV,>,?lHD6V@6[E,m,%)"<&(l1?)"&l0'/"&)1&lH&"0+=,
F[,>,?lHD6[@6VE,m,e%)"<&(l1?)"&l0'/"&)1&lH&"0+=,
4&50&-1,6&7010+',
!"#$%&'()*+',-+",8&50&-,"&7010+',
•  H*).*>+B*;./.01+01,(?&,H"+/&11,+-,/?)'#0'#,8&50&-1,
   (+,)=)H(,(?&,&H01(&%0/,1()(&,+-,)',)#&'(,(+,),'&C,
   H0&/&,+-,0'-+"%)*+'A,
•  IB5='*19&D01,01,/+'/&"'&=,C0(?,(?&,&7)5$)*+',
   +-,/5)0%1,8)1&=,+',H"&%01&1,0',+"=&",(+,"&)/?,
   /+'/5$10+'1A,
•  H09"+0>+9"*'>,,
   –  H"+70=&,8)10/,)'=,1$81()'*)5,(&/?'0G$&1,-+",(?&,)"(,+-,
      "&)1+'0'#@,)1,H&"-+"%&=,8I,?$%)',8&0'#1,0',&7&"I=)I,
      50-&,10($)*+'1@,
   –  #+,-)",8&I+'=,5+#0/)5,=&=$/*+'A,
!"#$%&'()*+',-+",8&50&-,"&7010+',
•  !#&'(,&'70"+'%&'(@,
•  4&50&-,"&7010+',=&1/"08&1,,
   –  (?&,C)I,0',C?0/?,)',)#&'(,01,1$HH+1&=,(+,/?)'#&,?&",
      8&50&-1,C?&','&C,0'-+"%)*+',)""07&1@,+",/?)'#&1,0',
      (?&,C+"5=,)"&,+81&"7&=f,,
•  !"#$%&'()*+',=&)51,,
   –  C0(?,1(")(&#0&1,)#&'(1,&%H5+I,-+",(?&0",+C',"&)1+'0'#@,,
   –  (+,/?)'#&,(?&,8&50&-1,+-,+(?&",)#&'(1@,8I,H"+70=0'#,
      "&)1+'1,-+",1$/?,/?)'#&A,
!"#$%&'()*+',-+",8&50&-,"&7010+',
   !"#$%&'()*+',)'=,8&50&-,"&7010+',)1,
        /+%H5&%&'()"I,=01/0H50'&1A,

 T)/?,'&&=1,(?&,+(?&"j1,1$HH+"(,(+,%+=&5,
  1$//&11-$5,=&/010+',%)<0'#,0',"&)5,C+"5=,
               )HH50/)*+'1A,
!"#$%&'()*+',_,4&50&-,6&7010+',
•  !',)"#$%&'(,!,-+",o,01,),1&(,+-,0'(&""&5)(&=,
   H0&/&1,+-,<'+C5&=#&,1$HH+"*'#,o,-"+%,
   &70=&'/&A,
•  F5)110/)5,8&50&-,"&7010+'>,`O&=,`'0(&,5)'#$)#&,p,
   C0(?,),/+%H5&(&,1&(,+-,8++5&)',/+''&/*7&1A,
•  ;)'I,=0J&"&'(,-")%&C+"<1,-+",8&50&-,"&7010+',
   C0(?,(?&0","&1H&/*7&,&H01(&%0/,%+=&51A,,
4&50&-,6&7010+',
•  $(./9*'.-+'04*)>,-+"%)501%,0',C?0/?,8&50&-1,)"&,
   "&H"&1&'(&=@,)'=,0',C?0/?,=0J&"&'(,<0'=1,+-,
   +H&")(+"1,/)',8&,=&`'&=A,,
•  4)10/,"&H"&1&'()*+',+-,&H01(&%0/,1()(&1>,
  –  7*).*>+/*9/>,1&(1,+-,1&'(&'/&1,/5+1&=,$'=&",5+#0/)5,
     /+'1&G$&'/&@,+",,
  –  7*).*>+7&/*/>,1&(1,+-,1&'(&'/&1,'+(,'&/&11)"05I,/5+1&=A,
  –  LH&")(+"1,H"&1&'(&=,0',(C+,C)I1>,,
     •  8I,#070'#,)',&OH50/0(,/+'1("$/*+',D)5#+"0(?%E,-+",(?&,
        +H&")(+"@,+",
     •  8I,#070'#,),1&(,+-,")*+')50(I,H+1($5)(&1,(+,8&,1)*1`&=,
        D/+'1(")0'(1EA,,
!"#$%&'()*+',_,4&50&-,6&7010+',
•    <*-*.;.15+1*!+.1>0B'&D01>,
      –  '&C,0'-+"%)*+',0',=0J&"&'(,1?)H&1,)'=,-+"%1@,&A#A@,3,01,),H"+H+10*+')5,-)/(f,
      –  8)10/,!g;,(?&+"I@,)11$%0'#,(?&,&H01(&%0/,1()(&,+-,(?&,)#&'(,(+,8&,#07&',8I,),8&50&-,1&(f,
      –  3,%0#?(,8&,%+"&,/+%H5&O@,&G$0HH&=,C0(?,=&#"&&,+-,H5)$108050(I@,+",?)7&,-+"%,+-,),"$5&@,+",),
         /+%H5&(&,)"#$%&'(@,+",+-,),1&(,+-,1$/?,&'**&1f,
•    $;&)=&D15+1*!+.1>0B'&D01>,,
      –  -+",-$"(?&",H"+/&110'#,3@,/"$/0)5,-+",(?&,)#&'(,(+,<'+C,0(1,+"0#0'@,)1,(?01,<'+C5&=#&,0'b$&'/&1,
         C0550'#'&11,(+,)=+H(,3f,
      –  3,8)1&=,+',)',+81&"7)*+',%)=&,8I,)#&'(,?&"1&5-@,1?&,C055,$1$)55I,8&,/+'70'/&=,+-,0(,8&0'#,("$&f,,
      –  0-,3,01,/+'7&I&=,8I,)'+(?&",)#&'(@,(?&,)#&'(,C055,"&G$0"&,1+%&,i$1*`/)*+',-+",3f,
      –  )1,%)'=)(+"I,1(&H,-+",")*+')5,(?0'<0'#@,1?&,&7)5$)(&,8+(?,3,)'=,H+11085&,i$1*`/)*+',+',(?&,
         8)101,+-,?&",+C',8&50&-1@,)'=,=&/0=&,0-,3,01,(+,8&,0'/+"H+")(&=,0'(+,?&",8&50&-1,+",'+(f,
•    +F"&15.15+7*).*>/>,,
      –  0-,(?&,)#&'(,=&/0=&=,(+,)=+H(,3@,1?&,&%H5+I1,1(")(&#0&1,(+,0'/+"H+")(&,3,/+'101(&'(5I,0'(+,?&",
         8&50&-1A,K+",(?01@,1?&,?)1,(+,$1&,8&50&-,"&7010+',(&/?'0G$&1,(+,/?)'#&,?&",&H01(&%0/,1()(&f,
•    J1>*B*1-*>,,
      –  K"+%,'&C,&H01(&%0/,1()(&@,(?&,)#&'(,=&"07&1,H5)$1085&,8&50&-1,(?)(,#$0=&,?&",8&?)70+$"f,
!"#$%&'()*+',_,4&50&-,6&7010+',
•  T%8&==0'#,0'(+,/+%H5&O,"&)1+'0'#,H"+/&11A,
•  9?&,/+%H5&%&'()"I,/?)")/(&"1,+-,)"#$%&'()*+',
   )'=,8&50&-,"&7010+',8&/+%&,&70=&'(>,
  –  )"#$%&'()*+',%)<&1,/+'("08$*+'1,(+,(?&,&7)5$)*+',
     1(&H@,
  –  8&50&-,"&7010+',01,&%H5+I&=,0',(?&,8&50&-,/?)'#&,H)"(A,,
–  T7)5$)*+',0'/5$=0'#,?IH+(?&*/)5,/?)'#&,
   H"+/&11&1>,
     •  C?)(,C+$5=,?)HH&',0-,(?&,'&C,0'-+"%)*+',C&"&,(+,8&,
        8&50&7&=@,,
     •  8&50&-,/?)'#&,0%H50/0(5I,"&50&1,+',5+#0/)5,50'<1,8&(C&&',H0&/&1,
        +-,0'-+"%)*+',C?0/?,/)',8&,"&H"&1&'(&=,8I,)"#$%&'(1A,,
!"#$%&'()*+',_,4&50&-,6&7010+',
•  4+(?,-"+%,)"#$%&'()*+',H"+/&11&1,)'=,-"+%,
   8&50&-,"&7010+',H"+/&11&1@,H5)$1085&,8&50&-1,/)',
   8&,+8()0'&=A,
•  4+(?,)"&)1,-+/$1,+'5I,+',H)"(1,+-,(?&,=I')%0/,
   "&)1+'0'#,H"+/&11,C?05&,)(,(?&,1)%&,*%&,
   H"+70=0'#,#&'&")5,)'=,7&"1)*5&,-")%&C+"<1A,,
F+%H)"01+',
•  ,.K*B*1-*/+7*9!**1+&B5='*19&D01+&14+7*).*>+B*;./.01>,
    –  "&H"&1&'()*+')5,011$&1@,1I'()/*/,)'=,1&%)'*/,-+$'=)*+'1,+-,8+(?,)"&)1f,,
    –  0',1()'=)"=,8&50&-,"&7010+'@,5+#0/)5,-+"%$5)1,$1&=,-+",M6@,"&1$5(1,+-,/?)'#&,
       H"+/&11&1,)"&,5+#0/)5,-+"%$5)1f,
    –  &H01(&%0/,1()(&1,)"&,/?)'#&=,C0(?,"&1H&/(,(+,&'("&'/?%&'(@,H5)$108050(I@,qf,
    –  (+,7&"0-I,"&1$5(1,+-,/?)'#&,H"+/&11&1@,/5)110/)5,5+#0/)5,1&%)'*/1,01,$1&=f,
    –  )"#$%&'()*+',-+/$1&1,+',0'(&")/*+'1,+-,)"#$%&'(1,)1,H0&/&1,+-,0'-+"%)*+',
       (?)(,%)I,)R)/<,+'&,)'+(?&"@,)'=,"&5)*+',8&(C&&',)"#$%&'(1,%)I,#07&,
       H"0+"0(I,(+,+'&,)"#$%&'(,+",)'+(?&"f,
    –  )"#$%&'(1,(?&%1&57&1,)"&,7&"I,?&(&"+#&'&+$1f,
    –  1&%)'*/1,%)<&1,H"&/01&,C?)(,#++=,)"#$%&'(1,)"&f,
•  F0''01+5B0=14/>,
    –  8+(?,=01/0H50'&1,)0%,)(,"&1+570'#,/+'b0/(1,C?0/?,)"&,$1$)55I,8)1&=,+',5+#0/)5,
       #"+$'=1@,0A&A@,+',/+'(")=0/*+'1f,
    –  %)<&,$1&,+-,H"&-&"&'/&,"&5)*+'1,(+,)/?0&7&,(?01,)0%f,
    –  8&50&-,"&7010+',H"+70=&1,),?0#?5I,=&/5)")*7&,-")%&C+"<,-+",(?)(@,8)1&=,+',
       H+1($5)(&1f,
    –  )"#$%&'()*+',01,%+"&,/+'/&"'&=,C0(?,H")/*/)5@,i$1*`/)*+'U8)1&=,
       (&/?'0G$&1f,
!"#$%&'()*+',0',4&50&-,6&7010+',
•  Q+C,)"#$%&'()*+',(&/?'0G$&1,/)',8&,$1&=,0',
   8&50&-,"&7010+',(?&+"IS,
  –  r$1*`/)*+'U8)1&=,("$(?,%)0'(&')'/&,1I1(&%1,s:+I5&t>,
     0'(&")/*+'1,8&(C&&',i$1*`/)*+'1,C?&',),'&C,
     i$1*`/)*+',?)1,8&&',)==&=@,(+,`'=,+$(,C?0/?,/+'/5$10+'1,
     /)',8&,i$1*`&=A,
  –  !11$%H*+'U8)1&=,("$(?,%)0'(&')'/&,1I1(&%1,sK)5)HH),&(,
     )5At>,%)')#0'#,)11$%H*+'1,0'1(&)=,+-,0%H5&%&'*'#,
     /?)'#&,H"+/&11&1@,/+%80'&,!9;.,0=&),C0(?,8)1&,"&7010+',
     )'=,1I1(&%,$10'#,)"#$%&'()*7&,1("$/($"&1,0',(?&,-+"%,+-,
     &OH5)')*+'1,-+","&7010+'1,+-,8&50&-,8)1&A,
4&50&-,6&7010+',0',!"#$%&'()*+',
•  Q+C,8&50&-,"&7010+',(&/?'0G$&1,/)',8&,$1&=,0',
   )"#$%&'()*+',(?&+"IS,
•  B+"<1,8I,s6+(1(&0't@,s;+#$055)'1<I,&(,)5At,)'=,s4+&55),&(,)5At,
   )%+'#,%+1(,/+%H"&?&'107&,)HH"+)/?&1,(+,)=="&11,),
   "&7010+',(?&+"I,-+",)"#$%&'(,1I1(&%1A,
•  .&7&")5,C)I1,+-,)HH5I0'#,8&50&-,"&7010+',0',)"#$%&'()*+'>,
   –    F?)'#0'#,8I,)==0'#,+",=&5&*'#,)',)"#$%&'(A,
   –    F?)'#0'#,8I,)==0'#,+",=&5&*'#,),1&(,+-,)"#$%&'(1A,
   –    F?)'#0'#,(?&,)R)/<,D)'=a+",=&-&)(E,"&5)*+',)%+'#,)"#$%&'(1A,
   –    F?)'#0'#,(?&,1()($1,+-,8&50&-1,D)1,/+'/5$10+'1,+-,)"#$%&'(1EA,
   –    F?)'#0'#,(?&,(IH&,+-,)',)"#$%&'(,D-"+%,1("0/(,(+,=&-&)1085&@,+",
        70/&,7&"1)EA,
9?"&&,)5(&"')*7&,)HH"+)/?&1,
•  K)5)HH)@,M&"'U318&"'&"@,)'=,.0%)"0A,<")("A&
   -"9(3(/$I&JF0)%$%2/$3&%$;&K"A"%3(H)"&
   -"%3/$($EG,!"*`/0)5,3'(&550#&'/&,r+$"')5@,
   VYV>Vu[^@,[[A,
•  F)I"+5@,=&,.)0'(,FI"@,)'=,p)#)1G$0&,./?0&OA,
   -"9(3(/$&/A&%$&.#E?5"$*%2/$&!L3*"5A,3',2"+/1,
   +-,M6,[^@,H)#&1,V[YuVXY@,[^A,
•  4&'-&"?)(@,:$8+01@,)'=,2")=&A,=/>&*/&($A"#&
   A#/5&($7/$3(3*"$*&H")("A3&>(*1/?*&#"9(3($EA,3',
   2"+/1,+-,3rF!3,V]]v@,H)#&1,VYY]uVYvv@,V]]vA,
K)5)HH),&(,)5A,)HH"+)/?,
•  !,<0'=,+-,'+'UH"0+"0*P&=,"&7010+',+H&")(+",
   8)1&=,+',(?&,$1&,+-,&OH5)')*+'1A,,
•  9?&,0=&),01,(?)(,)',)#&'(@,8&-+"&,0'/+"H+")*'#,
   0'-+"%)*+',0'/+'101(&'(,C0(?,0(1,<'+C5&=#&@,
   "&G$&1(1,)',&OH5)')*+',1$HH+"*'#,0(A,
•  F5)110/)5,=01*'/*+',8&(C&&'>,
  •  TOH5)')'=$%@,(?&,`')5,/+'/5$10+'@,
  •  TOH5)')'1@,1&(,+-,1&'(&'/&1,1$HH+"*'#,(?&,
     /+'/5$10+'A,
9?&,%+=&5,
•  .("$/($"&,+-,&OH5)')*+',10%05)",(+,1("$/($"&,+-,
   =&=$/*7&,)"#$%&'(A,
•  ,;)0',=0J&"&'/&>,,
   –  &7&"I,8&50&-,+-,)',&OH5)')*+',01,$'=&-&)1085&,D0',),%+%&'(,
      +-,*%&E@,
   –  1+%&,8&50&-1,+-,)',)"#$%&'(,%)I,8&,=&-&)1085&,+",
      (&'()*7&5I@,,
   –  &7&"I,&OH5)')*+',/+'()0'1,"$5&1,)'=,-)/($)5,<'+C5&=#&A,

  3-,(?&,1&'(&'/&1,0',(?&,&OH5)')'1,)"&,8&R&",+",%+"&,
  H5)$1085&,(?)',(?&,1&'(&'/&1,0',(?&,+"0#0')5,8&50&-,8)1&@,
           (?&',(?&,&OH5)')*+',01,0'/+"H+")(&=A,
9?&,%+=&5,
•  h+(,8&50&-1@,8$(,&OH5)')*+'1,D)'=,?&'/&,
   )"#$%&'(1E,1$HH+"*'#,),8&50&-,)"&,$1&=,-+",
   (?&,/?)'#&,H"+/&11A,,
•  F+'10=&",8+(?,<&"'&5,)'=,H)"*)5,%&&(,"&7010+',
   8I,),1&(,+-,1&'(&'/&1,)'=,#)7&,"&H"&1&'()*+',
   (?&+"&%1,-+",(?&%A,,
•  9?&1&,+H&")(+"1,%)I,H)"*)55I,)//&H(,(?&,'&C,
   0'-+"%)*+'@,1+,(?&I,)"&,'+'UH"0+"0*P&=A,
9?"&&,)5(&"')*7&,)HH"+)/?&1,
•  K)5)HH)@,M&"'U318&"'&"@,)'=,.0%)"0A,<")("A&
   -"9(3(/$I&JF0)%$%2/$3&%$;&K"A"%3(H)"&
   -"%3/$($EA,!"*`/0)5,3'(&550#&'/&,r+$"')5@,
   VYV>Vu[^@,[[A,
•  F)I"+5@,=&,.)0'(,FI"@,)'=,p)#)1G$0&,./?0&OA,
   -"9(3(/$&/A&%$&.#E?5"$*%2/$&!L3*"5A,3',2"+/1,
   +-,M6,[^@,H)#&1,V[YuVXY@,[^A,
•  4&'-&"?)(@,:$8+01@,)'=,2")=&A,=/>&*/&($A"#&
   A#/5&($7/$3(3*"$*&H")("A3&>(*1/?*&#"9(3($EA,3',
   2"+/1,+-,3rF!3,V]]v@,H)#&1,VYY]uVYvv@,V]]vA,
F)I"+5,&(,)5A,)HH"+)/?,
•  2"+H+1&,),:$'#U1(I5&,)81(")/(,)"#$%&'()*+',
   1I1(&%,)55+C0'#,(?&,)==0*+',+-,),'&C,
   )"#$%&'(,C?0/?,%)I,0'(&")/(,C0(?,H"&70+$1,
   )"#$%&'(1A,,
•  !"#$%&'()*+',-")%&C+"<,⟨!@6⟩,0=&'*`&=,
   C0(?,)',)11+/0)(&=,)R)/<,#")H?,gA,
•  6&7010+',H"+/&11,H"+=$/&1,),'&C,-")%&C+"<,
   "&H"&1&'(&=,8I,),#")H?,gj,)'=,),'&C,1&(,+-,
   &O(&'10+'1A,
9?&,%+=&5,
•  Q+C,(?&,1&(,+-,&O(&'10+'1,01,%+=0`&=,$'=&",(?&,"&7010+',
   H"+/&11S,
•  9IH+5+#I,+-,=0J&"&'(,"&7010+'1>,
   –  4*-./.;*+B*;./.01>,+'5I,+'&,)//&H()85&,1&(,+-,)"#$%&'(1,0',(?&,
      "&701&=,-")%&C+"<@,
   –  *%(&1/.;*+B*;./.01>,)==1,(?&,'&C,)"#$%&'(,(+,(?&,&O01*'#,
      &O(&'10+'1@,
   –  /*)*-D;*+B*;./.01>,=&/"&)1&,+-,(?&,'$%8&",+-,/?+0/&1@,
   –  L=*/D01.15+B*;./.01>,")01&,)%80#$0(I@,8I,0'/"&)10'#,(?&,'$%8&",
      +-,&O(&'10+'1@,
   –  4*/9B=-D;*+B*;./.01>,"&%+70'#,&7&"I,&O(&'10+'@,
   –  &)9*B.15+B*;./.01>,1+%&,&O(&'10+'1,D)55,+-,(?&%E,)"&,)5(&"&=A,
9?"&&,)5(&"')*7&,)HH"+)/?&1,
•  K)5)HH)@,M&"'U318&"'&"@,)'=,.0%)"0A,<")("A&
   -"9(3(/$I&JF0)%$%2/$3&%$;&K"A"%3(H)"&
   -"%3/$($EG,!"*`/0)5,3'(&550#&'/&,r+$"')5@,
   VYV>Vu[^@,[[A,
•  F)I"+5@,=&,.)0'(,FI"@,)'=,p)#)1G$0&,./?0&OA,
   -"9(3(/$&/A&%$&.#E?5"$*%2/$&!L3*"5G,3',2"+/1,
   +-,M6,[^@,H)#&1,V[YuVXY@,[^A,
•  4&'-&"?)(@,:$8+01@,)'=,2")=&A,=/>&*/&($A"#&
   A#/5&($7/$3(3*"$*&H")("A3&>(*1/?*&#"9(3($EG&3',
   2"+/1,+-,3rF!3,V]]v@,H)#&1,VYY]uVYvv@,V]]vA,
4&'-&"?)(,&(,)5A,)HH"+)/?,
•  ;+=&5,+"0&'(&=,(+C)"=1,(?&,("&)(%&'(,+-,
   0'/+'101(&'/I,/)$1&=,8I,(?&,$1&,+-,%$5*H5&,
   1+$"/&1,+-,0'-+"%)*+'A,,

•  M'+C5&=#&,8)1&1,)"&,1(")*`&=>,
  –  &)/?,-+"%$5),0',(?&,M4,01,)11+/0)(&=,C0(?,0(1,5&7&5,
     +-,/&"()0'(I,/+""&1H+'=0'#,(+,(?&,5)I&",(+,C?0/?,0(,
     8&5+'#1A,
9?&,%+=&5,
•  9C+,/5)11&1,+-,)HH"+)/?&1,(+,=&)5,C0(?,0'/+'101(&'/I,0',
   M4>,/+?&"&'/&,(?&+"0&1,)'=,-+$'=)*+',(?&+"0&1A,,
   –  K0"1(,0'101(1,+',"&7010'#,(?&,M4,)'=,"&1(+"0'#,/+'101(&'/I@,
   –  p)R&",)//&H(1,0'/+'101(&'/I,)'=,/+H&1,C0(?,0(A,
•  F+?&"&'/&,(?&+"0&1,H"+H+1&,,
   –  (+,#07&,$H,1+%&,-+"%$5)1,+-,(?&,M4,0',+"=&",(+,#&(,+'&,+",
      1&7&")5,/+'101(&'(,1$8U8)1&1@,
   –  (+,)HH5I,/5)110/)5,&'()05%&'(,+',(?&1&,/+'101(&'(,1$8U8)1&1,
      (+,=&=$/&,H5)$1085&,/+'/5$10+'1,+-,(?&,M4A,,
•  K+$'=)*+',(?&+"0&1,"&()0',,
   –  )55,)7)05)85&,0'-+"%)*+',,
   –  &)/?,H5)$1085&,/+'/5$10+',0'-&""&=,-"+%,(?&,M4,01,i$1*`&=,
      8I,1("+'#,)"#$%&'()*7&,"&)1+'1,-+",8&50&70'#,0',0(A,
9?&,%+=&5,
•  F5)0%>,,
   –  0(,=+&1,'+(,)5C)I1,%)<&,1&'1&,(+,"&701&,)',
      0'/+'101(&'(,M4@,0',H)"*/$5)"@,0-,0'-+"%)*+',/+%&1,
      -"+%,%$5*H5&,1+$"/&1f,
   •  '+(,&7&','&/&11)"I,(+,"&1(+"&,/+'101(&'/I,0',+"=&",
      (+,%)<&,1&'1085&,0'-&"&'/&1,-"+%,)',0'/+'101(&'(,
      M4@,10'/&,0'-&"&'/&,8)1&=,+',)"#$%&'()*+',/)',
      =&"07&,/+'/5$10+'1,)'=,"&)1+'1,(+,8&50&7&,(?&%@,
      0'=&H&'=&'(5I,+-,/+'101(&'/I,+-,(?&,M4f,
6&)1+'0'#,)8+$(,9"$1(,
9"$1(,0',;!.,
•  ;&/?)'01%,-+",%)')#0'#,$'/&"()0'(I,)8+$(,
   )$(+'+%+$1,&'**&1,)'=,(?&,0'-+"%)*+',(?&I,
   1(+"&A,
•  !#&'(1,?)7&,(+,"&)1+',)8+$(,,
  •  )%+$'(,(?)(,(?&I,("$1(,(?+1&,+(?&",&'**&1@,
  •  C?&(?&",(?&I,)"&,("$1*'#,(?+1&,&'**&1,(+,/)""I,
     +$(,1+%&,()1<@,+",,
  •  C?&(?&",(?&I,)"&,("$1*'#,(?+1&,&'**&1,(+,'+(,
     %01$1&,/"$/0)5,0'-+"%)*+'A,
9"$1(,
•  F)1(&5-")'/?0,)'=,K)5/+'&>,c%&5"$*%)&3*%*"I&%&
   7/50)"F&%M*?;"&/A&%$&%E"$*&F&*/>%#;3&
   %$/*1"#&%E"$*&L&%H/?*&*1"&H"1%9(/?#N%72/$&%&
   #")"9%$*&A/#&*1"&E/%)&EdA,

•  g)%8&R)>,c*#?3*&(3&*1"&3?HO"729"&0#/H%H()(*L&
   HL&>1(71&%$&($;(9(;?%)&.&"F0"7*3&*1%*&%$/*1"#&
   ($;(9(;?%)&<&0"#A/#53&%&E(9"$&%72/$&/$&>1(71&
   (*3&>")A%#"&;"0"$;3dA,,
9"$1(,
•  p0)$>,c(A&%E"$*&(&H")("9"3&*1%*&%E"$*&O&1%3&*/);&
   1(5&*1"&*#?*1&/$&0I&%$;&1"&*#?3*3&*1"&
   O?;E"5"$*&/A&O&/$&0I&*1"$&1"&>())&%)3/&H")("9"&
   0dA,,

•  F+%%+',&5&%&'(1,)"&,
   –  /+'101(&'(,=&#"&&,+-,$'/&"()0'(I@,)'=,,
   –  /+'b0/*'#,0'-+"%)*+',)11+/0)(&=,C0(?,("$1(A,,
!"#$%&'()*+',)'=,9"$1(,
•  M&I,)1H&/(,+-,("$1(>,0(,1(&%1,-"+%,(?&,
   "&5)*+'1?0H,8&(C&&',0'=070=$)51@,+",#"+$H1,+-,
   0'=070=$)51A,
•  9C+,%)i+",)1H&/(1,'&&=,(+,8&,?)'=5&=,8I,)'I,
   "&H"&1&'()*+',+-,("$1(>,
  –  %&)1$"&1,+-,("$1(@,)'=,
  –  H"+7&')'/&,+-,("$1(,0'-+"%)*+'A,
!"#$%&'()*+',)'=,9"$1(,
•  9"$1(,`&5=,!,%&(?+=+5+#0&1,-+",/+%H$*'#,
   ("$1(A,
•  !"#$%&'()*+',`&5=,!,%&(?+=+5+#0&1,-+",
   "&)1+'0'#,)8+$(,("$1(A,

  !"#$%&'()*+',/)',H"+70=&,),%&/?)'01%,-+",
  ?)'=50'#,%)'I,+-,(?&,)1H&/(1,(?)(,C&,'&&=,(+,
             /)H($"&,)8+$(,("$1(A,
!"#$%&'()*+',)'=,9"$1(,
•  !81(")/(,)HH"+)/?&1,1$/?,)1,:$'#,)'=,0(1,
   =&"07)*7&1@,("&)(,)"#$%&'(1,)1,)(+%0/,+8i&/(1A,
•  p0R5&,+",'+(?0'#,)8+$(,0'(&"')5,1("$/($"&,+-,
   )"#$%&'(A,
•  h+,%&/?)'01%,(+,"&H"&1&'(,(?&,1+$"/&,+-,(?&,
   0'-+"%)*+',-"+%,C?0/?,(?&,)"#$%&'(,01,
   /+'1("$/(&=A,
•  F)''+(,/)H($"&>,
  •  ),)R)/<1,8,8&/)$1&,8,01,8)1&=,+',0'-+"%)*+',-"+%,
     1+$"/&,1@,)'=,(?&"&,01,&70=&'/&,(?)(,1+$"/&,1,01,'+(,
     ("$1(C+"(?IA,
!"#$%&'()*+',)'=,9"$1(,
•  !'+(?&",H"+85&%,C0(?,:$'#j1,)"#$%&'()*+',
   1I1(&%,-"+%,(?&,H&"1H&/*7&,+-,"&)1+'0'#,
   )8+$(,("$1(,01,
  –  0(,?)1,'+,&OH50/0(,%&)'1,(+,"&H"&1&'(,=&#"&&1,+-,
     ("$1(A,,
•  F"0*/)5,H+0'(>,C?&(?&",1+%&+'&j1,)"#$%&'(,01,
   ("$1(C+"(?I,+",'+(@,(?&,H"&7)5&'/&,+-,
   '$%&"0/)5,%&)1$"&1,+-,("$1(,0',(?&,50(&")($"&,
   5&)=1,$1,(+,C)'(,(+,"&H"&1&'(,(?&1&A,
!"#$%&'()*+',)'=,9"$1(,
•  L(?&",0%H+"()'(,011$&1>,
  –  K&&=8)/<,-"+%,)"#$%&'(1,(+,1+$"/&1@,)'=,8)/<A,
  –  :0J&"&'(,H&"1H&/*7&1,+',("$1(>,,
     •  10'/&"0(I,k.,/+%H&(&'/&,
  –  :I')%0/1,+-,("$1(,)51+,=$&,(+,(?&,)R)/<1,H"&1&'(,
     0',(?&,)"#$%&'()*+',-")%&C+"<1A,
!"#$%&'()*+',)'=,9"$1(,
•  3%H+"()'(,)1H&/(,+-,"&)1+'0'#,)8+$(,("$1(>,
   '&&=,-+",)',0'=070=$)5,(+,8&,)85&,(+,"&701&,(?&,
   ("$1(,?&,?)1,0',)'+(?&",8)1&=,+',&OH&"0&'/&A,,
•  9?01,)55+C1,(+,"&H"&1&'(,(?&,/)1&,0',C?0/?,+'&,
   0'=070=$)5,"&701&1,0(1,70&C,+-,),1+$"/&,)1,),
   "&1$5(,+-,/+'10=&"0'#,0'-+"%)*+',H"+70=&=,8I,
   )'+(?&",0'=070=$)5A,
9?"&&,)5(&"')*7&,)HH"+)/?&1,
•  2)"1+'1@,9)'#@,.<5)"@,F)0@,;/4$"'&IA,
   .#E?5"$*%2/$PH%3";&#"%3/$($E&($&%E"$*3&>(*1&
   9%#L($E&;"E#""3&/A&*#?3*G&3',2"+/1,+-,!!;!.,[VVA,
•  =),F+1(),2&"&0")@,9&R)%)'P0@,k055)()A,F?)'#0'#,
   +'&j1,;0'=>,T")1&,+",6&C0'=S,2+11080501*/,4&50&-,
   6&7010+',C0(?,K$PPI,!"#$%&'()*+',4)1&=,+',
   9"$1(A,3',2"+/1,+-,3rF!3,[VV@,H)#&1,VZYUVWV@,
   [VVA,
•  ;)R@,;+"#&@,9+'0A,Q/5H($($E&3*%23273&%$;&
   %#E?5"$*3&*/&7/50?*"&*#?3*G,3',2"+/1,+-,!!;!.,
   [V@,H)#&1,[]U[VZ@,[VA,
2)"1+'1,&(,)5A,)HH"+)/?,
•  !11$%&,(?)(,)#&'(1,%)0'()0',),("$1(,'&(C+"<,+-,
   (?&0",)/G$)0'()'/&1@,C?0/?,0'/5$=&1,,
  –  ")*'#1,+-,?+C,%$/?,(?+1&,)/G$)0'()'/&1,)"&,("$1(&=@,
  –  ?+C,%$/?,(?+1&,)/G$)0'()'/&1,("$1(,(?&0",
     )/G$)0'()'/&1@,)'=,1+,+'A,
•  B?)(,0'-&"&'/&,01,"&)1+')85&,0',1$/?,'&(C+"<1S,
•  2"+H)#)*+',+-,("$1(,)'=,H"+7&')'/&,w,8+(?,(?&,
   (")'10*70(I,+-,("$1(,"&5)*+'1,)'=,%+"&,/+%H5&O,
   "&5)*+'1?0H1,50<&,c/+U/0()*+'jjA,
9?&,-")%&C+"<,
•  K")%&C+"<,(+,0'("+=$/&,(?&,1+$"/&1,0',)"#$%&'()*+'@,
   )'=,(+,&OH50/0(5I,&OH"&11,=&#"&&1,+-,("$1(A,,
•  !55,)#&'(1,(?)(,?)7&,/+%%$'0/)(&=,0'-+"%)*+',(+,!#,
   )"&,%&%8&"1,+-,!#j1,1+/0)5,'&(C+"<,
•  3(,01,H+11085&,,
   –  (+,/+'1("$/(,),#")H?,C?0/?,"&5)(&1,!#,(+,)55,(?&1&,)#&'(1f,,
   –  (+,)R)/?,),'$%&"0/)5,%&)1$"&,(+,&)/?,50'<,0',(?01,1+/0)5,
      '&(C+"<,(+,G$)'*-I,(?&,&O(&'(,(+,C?0/?,)',)#&'(,("$1(1,
      (?+1&,(+,C?0/?,0(,01,50'<&=,0',(?&,1+/0)5,'&(C+"<f,
•  .("$/($"&,/)55&=,("$1(,'&(C+"<A,
9?&,-")%&C+"<,
•  2"++-,("&&1,C?&"&,/+'/5$10+'1,+-,(?&,("&&1,)"&,
   -+"%$5)&,+-,0'(&"&1(A,
•  T)/?,-+"%$5),(?)(,01,'+(,)',0'-&"&'/&,01,50'<&=,(+,(?&,
   )#&'(,1$HH5I0'#,(?&,0'-+"%)*+'A,,
•  TA#A@,)"#$%&'(1,/+'/&"',C?&(?&",+",'+(,(+,C)(/?,
   !5%+=+7)",`5%,cQ)85&,/+',&55)dA,
•  4&50&-,7)5$&1,-+",0'-+"%)*+',(?)(,/+%&1,-"+%,=0J&"&'(,
   )#&'(1,1A(A,=&H&'=1,+',("$1(,0',(?&,1+$"/&,)#&'(D1EA,,
•  !"/1,0=&'*-I,/+'b0/(1,8&(C&&',)"#$%&'(1A,,
•  9?&1&,/+%H+'&'(1,%)<&,$H,),("$1(U&O(&'=&=,
   )"#$%&'()*+',#")H?A,
TO)%H5&,+-,("$1(,H"+H)#)*+',
TO)%H5&,
9?"&&,)5(&"')*7&,)HH"+)/?&1,
•  2)"1+'1@,9)'#@,.<5)"@,F)0@,;/4$"'&IA,
   .#E?5"$*%2/$PH%3";&#"%3/$($E&($&%E"$*3&>(*1&
   9%#L($E&;"E#""3&/A&*#?3*G&3',2"+/1,+-,!!;!.,[VVA,
•  =),F+1(),2&"&0")@,9&R)%)'P0@,k055)()A,F?)'#0'#,
   +'&j1,;0'=>,T")1&,+",6&C0'=S,2+11080501*/,4&50&-,
   6&7010+',C0(?,K$PPI,!"#$%&'()*+',4)1&=,+',
   9"$1(A,3',2"+/1,+-,3rF!3,[VV@,H)#&1,VZYUVWV@,
   [VVA,
•  ;)R@,;+"#&@,9+'0A,Q/5H($($E&3*%23273&%$;&
   %#E?5"$*3&*/&7/50?*"&*#?3*G,3',2"+/1,+-,!!;!.,
   [V@,H)#&1,[]U[VZ@,[VA,
=),F+1(),2&"&0"),&(,)5A,)HH"+)/?,
                                  ,
•  p+11,+-,0'-+"%)*+',0',/)1&,+-,"&0'1()(&%&'(,+-,
   H"&70+$1,0'-+"%)*+'A,

•  2"0'/0H5&,+-,cH"0+"0(I,(+,0'/+%0'#,0'-+"%)*+'d,
•  :")C8)/<1,0',%$5*)#&'(,1I1(&%1>,
  –  /?"+'+5+#0/)5,1&G$&'/&,+-,)""07)5,+-,0'-+"%)*+'@,
  –  '+(?0'#,(+,=+,C0(?,(?&0",("$1()8050(IA,
9?&,%+=&5
                          ,
•  !"#$%&'(1,⟨x@,y⟩,1$HH+"(,)#&'(1j,8&50&-1@,
•  4&50&-1,)"&,/+'/5$10+'1,+-,)"#$%&'(1@,
•  9"$1(C+"(?0'&11,%&)1$"&=,8I,$10'#,
   H"+8)8050*&1@4N9,+'5I,0-,=)(),)7)05)85&@,
•  2+11080501*/,5+#0/,C&55,1$0(&=,(+,=&)5,C0(?,
   0'/+%H5&(&,0'-+"%)*+'A,
9?&,%+=&5
                        ,
•  K$PPI,&7)5$)*+',+-,(?&,)"#$%&'(1@,,
•  !"#$%&'(1,)11+/0)(&=,C0(?,=&#"&&,+-,
   H5)$108050(I@,,
•  9"$1(C+"(?0'&11,+-,(?&,1+$"/&,+-,0'-+"%)*+',

•  !"#$%&'(1,&7)5$)(&=,0',#")=$)5,C)I,
   =&H&'=0'#,+',=&#"&&,+-,("$1(A,
9?&,%+=&5
                        ,

        '(#")*$

   +,-.,/' !

        !                   "
!"#$%       &       "   0       !
TO)%H5&,
                                                B
                                     I saw John killing Mary,
                                     thus John killed Mary.

                   A
                                                            Wit1
        If John did not kill Mary,
        then John is innocent.


Judge




                                                           D
                                            Mary was killed before 6 p.m.,
                                            thus when Mary was killed
                                            the show was still to begin

                   C
    John was at the theater with                                       Corner
    me when Mary was killed,
    thus John did not kill Mary.


 Wit2
TO)%H5&,
                                                  B                  A(B) = 0.2
                                       I saw John killing Mary,
A(A) = 1.0                             thus John killed Mary.

                     A
                                                              Wit1
          If John did not kill Mary,
          then John is innocent.


  Judge

                                                                        A(D) = 0.3
                                                             D
                                              Mary was killed before 6 p.m.,
                                              thus when Mary was killed
                                              the show was still to begin

                     C
      John was at the theater with                                       Corner
      me when Mary was killed,
      thus John did not kill Mary.


   Wit2                  A(C) = 1.0
TO)%H5&,
                                                  B                  α(B) = 0.2
                                       I saw John killing Mary,
α(A) = 0.8                             thus John killed Mary.

                     A
                                                              Wit1
          If John did not kill Mary,
          then John is innocent.


  Judge

                                                                        α(D) = 0.3
                                                             D
                                              Mary was killed before 6 p.m.,
                                              thus when Mary was killed
                                              the show was still to begin

                     C
      John was at the theater with                                       Corner
      me when Mary was killed,
      thus John did not kill Mary.


   Wit2                  α(C) = 0.7
9?"&&,)5(&"')*7&,)HH"+)/?&1,
•  2)"1+'1@,9)'#@,.<5)"@,F)0@,;/4$"'&IA,
   .#E?5"$*%2/$PH%3";&#"%3/$($E&($&%E"$*3&>(*1&
   9%#L($E&;"E#""3&/A&*#?3*G&3',2"+/1,+-,!!;!.,[VVA,
•  =),F+1(),2&"&0")@,9&R)%)'P0@,k055)()A,F?)'#0'#,
   +'&j1,;0'=>,T")1&,+",6&C0'=S,2+11080501*/,4&50&-,
   6&7010+',C0(?,K$PPI,!"#$%&'()*+',4)1&=,+',
   9"$1(A,3',2"+/1,+-,3rF!3,[VV@,H)#&1,VZYUVWV@,
   [VVA,
•  ;)R@,;+"#&@,9+'0A,Q/5H($($E&3*%23273&%$;&
   %#E?5"$*3&*/&7/50?*"&*#?3*G,3',2"+/1,+-,!!;!.,
   [V@,H)#&1,[]U[VZ@,[VA,
;)R,&(,)5A,)HH"+)/?,
•  .()"*'#,-"+%,z$,)'=,.0'#?>,)HH"+)/?,(+,("$1(,
   $10'#,:&%H1(&"U.?)-&",8&50&-,-$'/*+',=&"07&=,
   -"+%,1()*1*/)5,=)(),/+'/&"'0'#,()"#&(j1,
   8&?)70+$"A,
•  TO(&'10+',+-,z$,)'=,.0'#?j1,)HH"+)/?,8I,
   )55+C0'#,&7)5$)(+",(+,()<&,0'(+,)//+$'(@,0',
   )==0*+',(+,1()*1*/)5,=)()@,i$1*`&=,/5)0%1,
   /+'/&"'0'#,&OH&/(&=,8&?)70+$",+-,(?&,()"#&(A,
9?&,%+=&5,
•  F5)0%1,-+"%,8)101,+-,&7)5$)(+"j1,+H0'0+'1A,
•  K+"%)55I,"&H"&1&'(&=,8I,)"#$%&'(1,0',)81(")/(,
   )"#$%&'()*+'A,
•  9C+,/5)11&1,+-,)"#$%&'(1>,,
  –  >0B*-&/9+&B5='*19/@,0',-)7+$",+",)#)0'1(,("$1*'#,
     (?&,()"#&(@,)'=,,
  –  '.D5&D01+&B5='*19/@,)R)/<0'#,-+"&/)1(,
     )"#$%&'(1,+",+(?&",%0*#)*+',)"#$%&'(1A,
9?&,%+=&5,
•  ;&(?+=,-+",/+'1("$/*'#,:&%H1(&"U.?)-&",
   8&50&-,-$'/*+'1,-"+%,1()*1*/)5,=)(),)'=,(?&1&,
   )"#$%&'(1A,
•  h&C,)"#$%&'()*+'U8)1&=,8&50&-,-$'/*+',0',
   (&"%1,+-,'&C,)"#$%&'()*+'U8)1&=,&70=&'/&,
   %)11,-$'/*+',%l),/+%80'0'#,1()*1*/)5,
   &70=&'/&,)'=,)"#$%&'(1,)1,&70=&'/&A,
9?&,%+=&5,
•  K+",&)/?,=0%&'10+',=@,/+'10=&",)',)81(")/(,!K,K=,
   /+'101*'#,+-,(?&,-+55+C0'#,)"#$%&'(1>,
  –  K+"&/)1(,)"#$%&'(,e(,1$HH+"*'#,e9,D0A&A,{)De(E,|,e9E,
     +',(?&,#"+$'=,(?)(,(?&"&,01,'+,#$)")'(&&,0',(?&,-+"%,
     +-,C"0R&',/+'(")/(,/5)$1&,/+'/&"'0'#,=f,
  –  K+"&/)1(,)"#$%&'(,(,1$HH+"*'#,9,D0A&A,{),D(E|9E,+',(?&,
     #"+$'=,(?)(,(?&"&,&O01(1,),#$)")'(&&,0',(?&,-+"%,+-,),
     /+'(")/(,/5)$1&,/+'/&"'0'#,=f,
  –  ;0*#)*+',)"#$%&'(,7,)R)/<0'#,(,+',(?&,#"+$'=,(?)(,
     (?&,()"#&(,?)1,0',(?&,H)1(,c%+1(,+}&'d,70+5)(&=,
     &O01*'#,/+'(")/(,/5)$1&1,/+'/&"'0'#,=A,
9?&,%+=&5,
•  K+",&)/?,=0%&'10+',=@,,
   –  &0(?&",(?&"&,&O01(1,),/+'(")/(,/5)$1&,/+'/&"'0'#,=,)'=,
      (,8&5+'#1,(+,D1&(,+-,)"#$%&'(1,0'E,K=@,,
   –  +",(?&"&,=+&1,'+(,&O01(,1$/?,),/5)$1&@,)'=,e(,8&5+'#1,
      (+,K=A,,
•  3-,(?&"&,01,),/+'(")/(,/5)$1&@,)'=,0(,?)1,8&&',
   +81&"7&=,(?)(@,0',(?&,H)1(@,C0(?,)',)')5+#+$1,
   /5)$1&,(?&,0'(&")/*+',C0(?,()"#&(,=0=,'+(,&O?080(,
   =,)(,)//&H()85&,5&7&5,0',),%)i+"0(I,+-,/)1&1>,,
   –  K=,/+'()0'1,7,)1,C&55,)1,)',)R)/<,-"+%,7,)#)0'1(,(A,

More Related Content

Similar to T14 Argumentation for agent societies

Philosophy of religion
Philosophy of religionPhilosophy of religion
Philosophy of religionBraxton Hunter
 
Iron Chariots 2/2
Iron Chariots 2/2Iron Chariots 2/2
Iron Chariots 2/2Thoth Nine
 
Iron chariots arguments for the existence of flying spaghetti monster
Iron chariots arguments for the existence of flying spaghetti monsterIron chariots arguments for the existence of flying spaghetti monster
Iron chariots arguments for the existence of flying spaghetti monsterTriplo Sof
 
Averroes And The Teleological Argument
Averroes And The Teleological ArgumentAverroes And The Teleological Argument
Averroes And The Teleological ArgumentLuisa Polanco
 
2003_Does_God_Exist.ppt
2003_Does_God_Exist.ppt2003_Does_God_Exist.ppt
2003_Does_God_Exist.pptSoe Min Aung
 
The scientific method epistomy
The scientific method epistomyThe scientific method epistomy
The scientific method epistomyJulio Banks
 
Science Or Magic Which Should We Choose
Science Or Magic Which Should We ChooseScience Or Magic Which Should We Choose
Science Or Magic Which Should We ChooseDale Hull
 
Explanations 6.1 Qualities of Explanations Questions 0 of 3 com.docx
Explanations  6.1 Qualities of Explanations Questions 0 of 3 com.docxExplanations  6.1 Qualities of Explanations Questions 0 of 3 com.docx
Explanations 6.1 Qualities of Explanations Questions 0 of 3 com.docxgitagrimston
 
How to Design an Eternal Universe—Quantum Theory Implications
How to Design an Eternal Universe—Quantum Theory ImplicationsHow to Design an Eternal Universe—Quantum Theory Implications
How to Design an Eternal Universe—Quantum Theory ImplicationsGospel Conversations
 
Must A Darwinian be Skeptical About ReligionRachels Chapter .docx
Must A Darwinian be Skeptical About ReligionRachels Chapter .docxMust A Darwinian be Skeptical About ReligionRachels Chapter .docx
Must A Darwinian be Skeptical About ReligionRachels Chapter .docxrosemarybdodson23141
 
The big bang theory; The problem
The big bang theory; The problemThe big bang theory; The problem
The big bang theory; The problemMaikel
 
The shaky foundations of science slides - James Fodor
The shaky foundations of science slides - James FodorThe shaky foundations of science slides - James Fodor
The shaky foundations of science slides - James FodorAdam Ford
 
The big bang theory an analysis in the light of Quran
The big bang theory an analysis in the light of QuranThe big bang theory an analysis in the light of Quran
The big bang theory an analysis in the light of QuranSami Uddin
 
Science and Religion - Why is the world the way it is?
Science and Religion - Why is the world the way it is?Science and Religion - Why is the world the way it is?
Science and Religion - Why is the world the way it is?John Wilkins
 

Similar to T14 Argumentation for agent societies (20)

Philosophy of religion
Philosophy of religionPhilosophy of religion
Philosophy of religion
 
Iron Chariots 2/2
Iron Chariots 2/2Iron Chariots 2/2
Iron Chariots 2/2
 
Iron chariots arguments for the existence of flying spaghetti monster
Iron chariots arguments for the existence of flying spaghetti monsterIron chariots arguments for the existence of flying spaghetti monster
Iron chariots arguments for the existence of flying spaghetti monster
 
A2 Ontological
A2 OntologicalA2 Ontological
A2 Ontological
 
Philosophy04
Philosophy04Philosophy04
Philosophy04
 
Averroes And The Teleological Argument
Averroes And The Teleological ArgumentAverroes And The Teleological Argument
Averroes And The Teleological Argument
 
As Teleological
As TeleologicalAs Teleological
As Teleological
 
2003_Does_God_Exist.ppt
2003_Does_God_Exist.ppt2003_Does_God_Exist.ppt
2003_Does_God_Exist.ppt
 
Large Hadron Collider
Large Hadron ColliderLarge Hadron Collider
Large Hadron Collider
 
The scientific method epistomy
The scientific method epistomyThe scientific method epistomy
The scientific method epistomy
 
Science Or Magic Which Should We Choose
Science Or Magic Which Should We ChooseScience Or Magic Which Should We Choose
Science Or Magic Which Should We Choose
 
Explanations 6.1 Qualities of Explanations Questions 0 of 3 com.docx
Explanations  6.1 Qualities of Explanations Questions 0 of 3 com.docxExplanations  6.1 Qualities of Explanations Questions 0 of 3 com.docx
Explanations 6.1 Qualities of Explanations Questions 0 of 3 com.docx
 
How to Design an Eternal Universe—Quantum Theory Implications
How to Design an Eternal Universe—Quantum Theory ImplicationsHow to Design an Eternal Universe—Quantum Theory Implications
How to Design an Eternal Universe—Quantum Theory Implications
 
The "God Particle," Dark Matter, Black Holes, and All That
The "God Particle," Dark Matter, Black Holes, and All ThatThe "God Particle," Dark Matter, Black Holes, and All That
The "God Particle," Dark Matter, Black Holes, and All That
 
Must A Darwinian be Skeptical About ReligionRachels Chapter .docx
Must A Darwinian be Skeptical About ReligionRachels Chapter .docxMust A Darwinian be Skeptical About ReligionRachels Chapter .docx
Must A Darwinian be Skeptical About ReligionRachels Chapter .docx
 
The big bang theory; The problem
The big bang theory; The problemThe big bang theory; The problem
The big bang theory; The problem
 
The shaky foundations of science slides - James Fodor
The shaky foundations of science slides - James FodorThe shaky foundations of science slides - James Fodor
The shaky foundations of science slides - James Fodor
 
A2 Cosmological
A2 CosmologicalA2 Cosmological
A2 Cosmological
 
The big bang theory an analysis in the light of Quran
The big bang theory an analysis in the light of QuranThe big bang theory an analysis in the light of Quran
The big bang theory an analysis in the light of Quran
 
Science and Religion - Why is the world the way it is?
Science and Religion - Why is the world the way it is?Science and Religion - Why is the world the way it is?
Science and Religion - Why is the world the way it is?
 

More from EASSS 2012

T12 Distributed search and constraint handling
T12	Distributed search and constraint handlingT12	Distributed search and constraint handling
T12 Distributed search and constraint handlingEASSS 2012
 
T4 Introduction to the modelling and verification of, and reasoning about mul...
T4 Introduction to the modelling and verification of, and reasoning about mul...T4 Introduction to the modelling and verification of, and reasoning about mul...
T4 Introduction to the modelling and verification of, and reasoning about mul...EASSS 2012
 
T13 Agent coordination in planning and scheduling
T13	Agent coordination in planning and schedulingT13	Agent coordination in planning and scheduling
T13 Agent coordination in planning and schedulingEASSS 2012
 
T3 Agent oriented programming languages
T3 Agent oriented programming languagesT3 Agent oriented programming languages
T3 Agent oriented programming languagesEASSS 2012
 
T0. Multiagent Systems and Electronic Institutions
T0. Multiagent Systems and Electronic InstitutionsT0. Multiagent Systems and Electronic Institutions
T0. Multiagent Systems and Electronic InstitutionsEASSS 2012
 
T2. Organization and Environment oriented programming
T2. Organization and Environment oriented programmingT2. Organization and Environment oriented programming
T2. Organization and Environment oriented programmingEASSS 2012
 
T11. Normative multi-agent systems
T11. Normative multi-agent systemsT11. Normative multi-agent systems
T11. Normative multi-agent systemsEASSS 2012
 
T9. Trust and reputation in multi-agent systems
T9. Trust and reputation in multi-agent systemsT9. Trust and reputation in multi-agent systems
T9. Trust and reputation in multi-agent systemsEASSS 2012
 

More from EASSS 2012 (8)

T12 Distributed search and constraint handling
T12	Distributed search and constraint handlingT12	Distributed search and constraint handling
T12 Distributed search and constraint handling
 
T4 Introduction to the modelling and verification of, and reasoning about mul...
T4 Introduction to the modelling and verification of, and reasoning about mul...T4 Introduction to the modelling and verification of, and reasoning about mul...
T4 Introduction to the modelling and verification of, and reasoning about mul...
 
T13 Agent coordination in planning and scheduling
T13	Agent coordination in planning and schedulingT13	Agent coordination in planning and scheduling
T13 Agent coordination in planning and scheduling
 
T3 Agent oriented programming languages
T3 Agent oriented programming languagesT3 Agent oriented programming languages
T3 Agent oriented programming languages
 
T0. Multiagent Systems and Electronic Institutions
T0. Multiagent Systems and Electronic InstitutionsT0. Multiagent Systems and Electronic Institutions
T0. Multiagent Systems and Electronic Institutions
 
T2. Organization and Environment oriented programming
T2. Organization and Environment oriented programmingT2. Organization and Environment oriented programming
T2. Organization and Environment oriented programming
 
T11. Normative multi-agent systems
T11. Normative multi-agent systemsT11. Normative multi-agent systems
T11. Normative multi-agent systems
 
T9. Trust and reputation in multi-agent systems
T9. Trust and reputation in multi-agent systemsT9. Trust and reputation in multi-agent systems
T9. Trust and reputation in multi-agent systems
 

Recently uploaded

Textual Evidence in Reading and Writing of SHS
Textual Evidence in Reading and Writing of SHSTextual Evidence in Reading and Writing of SHS
Textual Evidence in Reading and Writing of SHSMae Pangan
 
Grade 9 Quarter 4 Dll Grade 9 Quarter 4 DLL.pdf
Grade 9 Quarter 4 Dll Grade 9 Quarter 4 DLL.pdfGrade 9 Quarter 4 Dll Grade 9 Quarter 4 DLL.pdf
Grade 9 Quarter 4 Dll Grade 9 Quarter 4 DLL.pdfJemuel Francisco
 
ICS2208 Lecture6 Notes for SL spaces.pdf
ICS2208 Lecture6 Notes for SL spaces.pdfICS2208 Lecture6 Notes for SL spaces.pdf
ICS2208 Lecture6 Notes for SL spaces.pdfVanessa Camilleri
 
Oppenheimer Film Discussion for Philosophy and Film
Oppenheimer Film Discussion for Philosophy and FilmOppenheimer Film Discussion for Philosophy and Film
Oppenheimer Film Discussion for Philosophy and FilmStan Meyer
 
Student Profile Sample - We help schools to connect the data they have, with ...
Student Profile Sample - We help schools to connect the data they have, with ...Student Profile Sample - We help schools to connect the data they have, with ...
Student Profile Sample - We help schools to connect the data they have, with ...Seán Kennedy
 
ClimART Action | eTwinning Project
ClimART Action    |    eTwinning ProjectClimART Action    |    eTwinning Project
ClimART Action | eTwinning Projectjordimapav
 
week 1 cookery 8 fourth - quarter .pptx
week 1 cookery 8  fourth  -  quarter .pptxweek 1 cookery 8  fourth  -  quarter .pptx
week 1 cookery 8 fourth - quarter .pptxJonalynLegaspi2
 
Man or Manufactured_ Redefining Humanity Through Biopunk Narratives.pptx
Man or Manufactured_ Redefining Humanity Through Biopunk Narratives.pptxMan or Manufactured_ Redefining Humanity Through Biopunk Narratives.pptx
Man or Manufactured_ Redefining Humanity Through Biopunk Narratives.pptxDhatriParmar
 
Visit to a blind student's school🧑‍🦯🧑‍🦯(community medicine)
Visit to a blind student's school🧑‍🦯🧑‍🦯(community medicine)Visit to a blind student's school🧑‍🦯🧑‍🦯(community medicine)
Visit to a blind student's school🧑‍🦯🧑‍🦯(community medicine)lakshayb543
 
How to Fix XML SyntaxError in Odoo the 17
How to Fix XML SyntaxError in Odoo the 17How to Fix XML SyntaxError in Odoo the 17
How to Fix XML SyntaxError in Odoo the 17Celine George
 
Reading and Writing Skills 11 quarter 4 melc 1
Reading and Writing Skills 11 quarter 4 melc 1Reading and Writing Skills 11 quarter 4 melc 1
Reading and Writing Skills 11 quarter 4 melc 1GloryAnnCastre1
 
MS4 level being good citizen -imperative- (1) (1).pdf
MS4 level   being good citizen -imperative- (1) (1).pdfMS4 level   being good citizen -imperative- (1) (1).pdf
MS4 level being good citizen -imperative- (1) (1).pdfMr Bounab Samir
 
Scientific Writing :Research Discourse
Scientific  Writing :Research  DiscourseScientific  Writing :Research  Discourse
Scientific Writing :Research DiscourseAnita GoswamiGiri
 
Unraveling Hypertext_ Analyzing Postmodern Elements in Literature.pptx
Unraveling Hypertext_ Analyzing  Postmodern Elements in  Literature.pptxUnraveling Hypertext_ Analyzing  Postmodern Elements in  Literature.pptx
Unraveling Hypertext_ Analyzing Postmodern Elements in Literature.pptxDhatriParmar
 
Grade Three -ELLNA-REVIEWER-ENGLISH.pptx
Grade Three -ELLNA-REVIEWER-ENGLISH.pptxGrade Three -ELLNA-REVIEWER-ENGLISH.pptx
Grade Three -ELLNA-REVIEWER-ENGLISH.pptxkarenfajardo43
 
Multi Domain Alias In the Odoo 17 ERP Module
Multi Domain Alias In the Odoo 17 ERP ModuleMulti Domain Alias In the Odoo 17 ERP Module
Multi Domain Alias In the Odoo 17 ERP ModuleCeline George
 

Recently uploaded (20)

Textual Evidence in Reading and Writing of SHS
Textual Evidence in Reading and Writing of SHSTextual Evidence in Reading and Writing of SHS
Textual Evidence in Reading and Writing of SHS
 
Grade 9 Quarter 4 Dll Grade 9 Quarter 4 DLL.pdf
Grade 9 Quarter 4 Dll Grade 9 Quarter 4 DLL.pdfGrade 9 Quarter 4 Dll Grade 9 Quarter 4 DLL.pdf
Grade 9 Quarter 4 Dll Grade 9 Quarter 4 DLL.pdf
 
ICS2208 Lecture6 Notes for SL spaces.pdf
ICS2208 Lecture6 Notes for SL spaces.pdfICS2208 Lecture6 Notes for SL spaces.pdf
ICS2208 Lecture6 Notes for SL spaces.pdf
 
Oppenheimer Film Discussion for Philosophy and Film
Oppenheimer Film Discussion for Philosophy and FilmOppenheimer Film Discussion for Philosophy and Film
Oppenheimer Film Discussion for Philosophy and Film
 
Student Profile Sample - We help schools to connect the data they have, with ...
Student Profile Sample - We help schools to connect the data they have, with ...Student Profile Sample - We help schools to connect the data they have, with ...
Student Profile Sample - We help schools to connect the data they have, with ...
 
ClimART Action | eTwinning Project
ClimART Action    |    eTwinning ProjectClimART Action    |    eTwinning Project
ClimART Action | eTwinning Project
 
week 1 cookery 8 fourth - quarter .pptx
week 1 cookery 8  fourth  -  quarter .pptxweek 1 cookery 8  fourth  -  quarter .pptx
week 1 cookery 8 fourth - quarter .pptx
 
Man or Manufactured_ Redefining Humanity Through Biopunk Narratives.pptx
Man or Manufactured_ Redefining Humanity Through Biopunk Narratives.pptxMan or Manufactured_ Redefining Humanity Through Biopunk Narratives.pptx
Man or Manufactured_ Redefining Humanity Through Biopunk Narratives.pptx
 
prashanth updated resume 2024 for Teaching Profession
prashanth updated resume 2024 for Teaching Professionprashanth updated resume 2024 for Teaching Profession
prashanth updated resume 2024 for Teaching Profession
 
Faculty Profile prashantha K EEE dept Sri Sairam college of Engineering
Faculty Profile prashantha K EEE dept Sri Sairam college of EngineeringFaculty Profile prashantha K EEE dept Sri Sairam college of Engineering
Faculty Profile prashantha K EEE dept Sri Sairam college of Engineering
 
Visit to a blind student's school🧑‍🦯🧑‍🦯(community medicine)
Visit to a blind student's school🧑‍🦯🧑‍🦯(community medicine)Visit to a blind student's school🧑‍🦯🧑‍🦯(community medicine)
Visit to a blind student's school🧑‍🦯🧑‍🦯(community medicine)
 
INCLUSIVE EDUCATION PRACTICES FOR TEACHERS AND TRAINERS.pptx
INCLUSIVE EDUCATION PRACTICES FOR TEACHERS AND TRAINERS.pptxINCLUSIVE EDUCATION PRACTICES FOR TEACHERS AND TRAINERS.pptx
INCLUSIVE EDUCATION PRACTICES FOR TEACHERS AND TRAINERS.pptx
 
Mattingly "AI & Prompt Design: Large Language Models"
Mattingly "AI & Prompt Design: Large Language Models"Mattingly "AI & Prompt Design: Large Language Models"
Mattingly "AI & Prompt Design: Large Language Models"
 
How to Fix XML SyntaxError in Odoo the 17
How to Fix XML SyntaxError in Odoo the 17How to Fix XML SyntaxError in Odoo the 17
How to Fix XML SyntaxError in Odoo the 17
 
Reading and Writing Skills 11 quarter 4 melc 1
Reading and Writing Skills 11 quarter 4 melc 1Reading and Writing Skills 11 quarter 4 melc 1
Reading and Writing Skills 11 quarter 4 melc 1
 
MS4 level being good citizen -imperative- (1) (1).pdf
MS4 level   being good citizen -imperative- (1) (1).pdfMS4 level   being good citizen -imperative- (1) (1).pdf
MS4 level being good citizen -imperative- (1) (1).pdf
 
Scientific Writing :Research Discourse
Scientific  Writing :Research  DiscourseScientific  Writing :Research  Discourse
Scientific Writing :Research Discourse
 
Unraveling Hypertext_ Analyzing Postmodern Elements in Literature.pptx
Unraveling Hypertext_ Analyzing  Postmodern Elements in  Literature.pptxUnraveling Hypertext_ Analyzing  Postmodern Elements in  Literature.pptx
Unraveling Hypertext_ Analyzing Postmodern Elements in Literature.pptx
 
Grade Three -ELLNA-REVIEWER-ENGLISH.pptx
Grade Three -ELLNA-REVIEWER-ENGLISH.pptxGrade Three -ELLNA-REVIEWER-ENGLISH.pptx
Grade Three -ELLNA-REVIEWER-ENGLISH.pptx
 
Multi Domain Alias In the Odoo 17 ERP Module
Multi Domain Alias In the Odoo 17 ERP ModuleMulti Domain Alias In the Odoo 17 ERP Module
Multi Domain Alias In the Odoo 17 ERP Module
 

T14 Argumentation for agent societies

  • 1. Argumentation for Agent Societies Part I 4%33(5()(%$/&6(%7/5($& !"#"$%&'())%*%& 8$(9"#3(*%:&;(&<#"37(%& +,-+.&!/01(%&.$20/)(3& 1
  • 2. Introduction to the tutorial Argumentation for Agent Societies (Some) answers to the following two questions: 1.  What’s argumentation? ! mainly today 2.  What is argumentation good for (in the MAS context)? ! mainly tomorrow Let’s start with the first question… 2
  • 3. An informal example (1) We should run Large Hadron Collider The conclusion LHC allows us to Understanding understand the Laws the Laws of the The reason of the Universe Universe is good We are justified in believing that we should run LHC ! 3
  • 4. An informal example (2) We should run Large Hadron Collider The conclusion LHC allows us to Understanding understand the Laws the Laws of the The reason of the Universe Universe is good We are justified in believing that we should run LHC ! BUT In Argumentation (and in real life as well): - reasons are not necessary “conclusive” (they don’t logically entail conclusions) - arguments and conclusions can be “retracted” in front of new information, i.e. counterarguments 4
  • 5. An informal example (3) We should run Large Hadron Collider We should not run LHC LHC allows us to Understanding LHC will generate Destroying understand the Laws the Laws of the black holes Earth of the Universe Universe is good destroying Earth is bad Now we are justified in believing that we should not run LHC " 5
  • 6. An informal example (4) We should run Large Hadron Collider We should not run LHC LHC allows us to Understanding LHC will generate Destroying understand the Laws the Laws of the black holes Earth of the Universe Universe is good destroying Earth is bad Black holes will not destroy Earth Black holes will evaporate because of Hawking radiation Now we are again justified in believing that we should run LHC ! 6
  • 7. An informal example (5) We should run Large Hadron Collider We should not run LHC LHC allows us to Understanding LHC will generate Destroying understand the Laws the Laws of the black holes Earth of the Universe Universe is good destroying Earth is bad Hawking radiation Black holes will does not exist not destroy Earth Black holes will Dr Azzeccagarbugli evaporate because says so of Hawking radiation Now we are again justified in believing that we should not run LHC " 7
  • 8. An informal example (6) We should run Large Hadron Collider We should not run LHC LHC allows us to Understanding LHC will generate Destroying understand the Laws the Laws of the black holes Earth of the Universe Universe is good destroying Earth is bad Hawking radiation Black holes will does not exist not destroy Earth Black holes will Dr Azzeccagarbugli evaporate because says so of Hawking radiation Dr Azzeccagarbugli Now we are again justified is not expert in physics in believing that we should He is a lawyer run LHC ! 8
  • 9. What s argumentation? (1) [Prakken 2011] Argumentation is the process of supporting claims with grounds and defending them against attack. [van Eemeren et al, 1996] Argumentation is a verbal and social activity of reason aimed at increasing (or decreasing) the acceptability of a controversial standpoint for the listener or reader, by putting forward a constellation of propositions intended to justify (or refute) the standpoint before a rational judge. •  A framework for practical and uncertain reasoning able to cope with partial and inconsistent knowledge - philosophical roots: Aristotle, Toulmin (1958) - in AI: R.P. Loui (1987), J. Pollock (1987), G. Simari & Loui (1992) 9
  • 10. What s argumentation? (2) The elements of an argumentation system •  The definition of argument (possibly including an underlying logical language + a notion of logical consequence) •  The notion of attack and defeat (successful attack) between arguments •  An argumentation semantics selecting acceptable (justified) arguments 10
  • 11. Definition of argument: several possibilities (1) •  ASSUMPTION-BASED ARGUMENTATION Given a knowledge base (K, Ass) Consistent theory Set of assumptions ARGUMENT for p: (A, p) such that - A " Ass - A # K is consistent and entails p - There is no A’$A such that A’ # K entails p ATTACKS to an argument: on its assumptions [see Besnard&Hunter, Dung-Kowalski-Toni] 11
  • 12. Definition of argument: several possibilities (2) •  ARGUMENT SCHEMES - correspond to recurring patterns of reasoning - have associated “critical questions” Example: Expert Testimony [WALTON 1996] E is expert on D E says P P is in D Therefore, P is the case Critical questions: Is E biased? Is P consistent with what other experts say? Is P consistent with known evidence? 12
  • 13. Definition of argument: several possibilities (3) •  ARGUMENT SCHEMES IN A MEDICAL APPLICATION Viability Scheme Organ O of donor D is available No contraindications are known for donating O to recipient R Therefore, organ O is viable CRITICAL QUESTIONS: Does donor D have a contraindication for donating organ O? Nonviability Scheme Donor D of organ O has condition C C is a contraindication are for donating O Therefore, organ O is nonviable [Tolchinsky et al, 2006] 13
  • 14. Definition of argument: several possibilities (4) •  STABLE MARRIAGE PROBLEM - Arguments of the kind <Alice, John> - <Barbara, John> attacks <Alice, John> if John prefers Barbara to Alice •  PLANNING - Plans as arguments (that a goal will be achieved) - Defeat between plans as attacks ……… In general Arguments take different forms (domain-independent vs. domain dependent) Today examples will refer to rule-based approaches… 14
  • 15. Rule-based approaches •  ARGUMENT a tree made up of rules of inference constructed from a set of premises to reach a conclusion •  Two kinds of rules: A (0.7) % ¬C (0.7)   A % B: deductive - indefeasible B (0.9)   A ! B: non-deductive - defeasible D (0.9) ! C (0.8) •  A strength value may be associated to premises and rules, giving rise to argument strength See [J.Pollock, 1992], [G. Vreeswijk, 1997], … 15
  • 16. Rule-based approaches (2) Notion of conflict –  Rebutting: an argument attacks another one by denying its [possibly intermediate] conclusion –  Undercutting: an argument attacks the applicability of a defeasible rule of inference Notion of defeat A An argument ' defeats ( iff: % ¬C B - ' undercuts (, or - ' rebuts ( and D!C ' is not weaker than ( E!(D&C) [Pollock 92] 16
  • 17. Rule-based approaches (3) EXAMPLE REBUTTING DEFEAT UNDERCUTTING It’s It’s not DEFEAT Bob raining raining is unreliable Smith says Bob says Bob is drunk it’s raining it’s not raining 17
  • 18. The ASPIC framework •  One result of the European ASPIC Project (2004-2006) •  Generalizes Pollock’s rule-based approach in several respects: - any logical language (and an associated ‘contrariness function’ generalizing classical negation) can be adopted - can be instantiated by a partial preorder on defeasible rules - premises are distinguished into necessary, ordinary and assumption premises (ordinary and assumption premises partially preordered) - a partial preorder is assumed between arguments •  Besnard & Hunter’s approach, Pollock’s system… can be obtained as instances of ASPIC framework •  See [H. Prakken, “An abstract framework for argumentation with structured arguments”, Argument and Computation, 2010] for details. 18
  • 19. Argumentation in the context of MAS (1) Advantageous features •  Several kinds of arguments can be represented - epistemic reasoning - practical reasoning •  Able to handle uncertain and partial knowledge - nonmonotonic notion of warrant: 1) wrt further information 2) wrt further reasoning steps (anytime reasoning framework) •  A natural representation + justification of choices (in terms of argument , rebuttal , counterargument …) •  Argumentation has a dialogical side (in terms of argument , attack , defence …) 19
  • 20. Argumentation in the context of MAS (2) The uses of argumentation (examples) AUTONOMOUS MULTI-AGENT REASONING INTERACTION EPISTEMIC - Belief Revision REASONING (arguing over beliefs) - Trust management (arguing over other - Dialectics in agents reputation) Multiagent Interaction PRATICAL - Decision making REASONING (arguing about the expected value of possible actions) 20
  • 21. What s argumentation? (3) The elements of an argumentation system •  The definition of argument (possibly including an underlying logical language + a notion of logical consequence) •  The notion of attack and defeat (successful attack) between arguments •  An argumentation semantics selecting acceptable (justified) arguments 21
  • 22. What s abstract argumentation? Usually “abstract” stands for a difficult thing… Here it means “simple”! The elements of an argumentation system •  The definition of argument (possibly including an underlying logical language + a notion of logical consequence) •  The notion of attack and defeat (successful attack) between arguments •  An argumentation semantics selecting acceptable (justified) arguments Abstract argumentation focuses on this aspect 22
  • 23. Dung s argumentation framework [Dung ’95] AF = <A, %> attack (or defeat) relation [unspecified definition] Arguments [origin and structure not specified] •  Graphical representation as a directed graph [defeat graph], e.g. Representation of LHC example Representation of weather example 23
  • 24. Dung s argumentation framework (2) So, what remains to be done? ARGUMENT EVALUATION: GIVEN AN ARGUMENTATION FRAMEWORK, DETERMINE THE JUSTIFICATION STATE (ALSO CALLED DEFEAT STATUS) OF ARGUMENTS, IN PARTICULAR: WHAT ARGUMENTS EMERGE UNDEFEATED FROM THE CONFLICT, I.E. ARE ACCEPTABLE? 24
  • 25. Argumentation semantics •  Specification of a method for argument evaluation, or of criteria to determine, given a set of arguments, their defeat status Semantics Argumentation Framework Defeat status Undefeated Defeat status Defeated Provisionally Defeated 25
  • 26. Labelling vs. extension-based semantics LABELLING-BASED SEMANTICS - Based on the notion of labelling [assignment to each argument of a label from a predefined set] - Specifies how to derive from an argumentation framework a set of labellings - Justification of arguments derived from the set of labellings EXTENSION-BASED SEMANTICS - Less general (at least in theory), but more common kind of semantics -  Based on the notion of extension [set of arguments collectively acceptable ] 26
  • 27. Extension-based semantics Semantics S Argumentation framework AF Set of extensions S(AF) 27
  • 28. From extensions to defeat status Set of extensions S(AF) Defeat/Justification Status A common definition •  Skeptically justified argument: belongs to all of the extensions •  Credulously justified argument: belongs to at least one •  Indefensible argument: does not belong to any extension 28
  • 29. Unique-status vs. multiple-status semantics Unique-Status Semantics Unique extension: empty set ') () ' and ( directly unjustified (provisionally defeated) ') () Multiple-Status Semantics ') () ') () ! ' and ( unjustified (provisionally defeated) 29
  • 30. Relationship between labelling and extension-based approaches •  Almost all approaches adopt the set {IN, OUT, UNDEC} - IN = belonging to the extension - OUT = attacked by the extension - UNDEC= not belonging to nor attacked by the extension Unique-Status Semantics ') () UNDEC UNDEC ') () Multiple-Status Semantics ') () ') () IN OUT OUT IN 30
  • 31. The core of Dung’s theory: complete “semantics” Acceptability ' acceptable w.r.t. (“defended by”) S ' •  all attackers of ' are attacked by S Admissible set S S •  conflict-free •  every element acceptable w.r.t. S (defends all of its elements) Complete semantics Complete IF extension also includes all acceptable elements w.r.t. itself All traditional semantics select complete extensions 31
  • 32. Complete “semantics”: examples Chain Admissible sets: ø, {'}, {', *} ') () *) Only one complete extension: CO(AF) = {{', *}} Nixon Diamond ') () All admissible sets are complete ') () CO(AF) = ') () { ø, {'}, {(} } ') () 32
  • 33. Complete “semantics”: examples (2) Nixon Diamond + node Admissible sets: ') () *) ø, {'}, {(}, {', *} CO(AF) = { ') () *) ø CO(AF) ') () *) {', *}, ') () *) {(} } 33
  • 34. The Grounded Semantics: a unique status approach Grounded extension GE(AF): Least complete extension included in all extensions of any traditional semantics Grounded semantics is the “most skeptical” one Undefeated Defeat status Defeated Provisionally Defeated 34
  • 35. Grounded semantics: examples Chain ') () *) GE(AF) = {', *} Nixon Diamond ') () GE(AF) = ø Nixon Diamond + node ') () *) GE(AF) = ø 35
  • 36. Floating arguments: a problem for grounded semantics •  Actually, grounded semantics is polynomially computable •  But sometimes a more discriminative behavior is desirable THE CASE OF FLOATING ARGUMENTS () () *) +) VS *) +) ') ') Grounded Semantics What we (may) want •  A problem for all possible unique status approaches Let us consider multiple status approaches! 36
  • 37. Stable Semantics Stable extension = conflict-free set attacking all outside arguments THE CASE OF FLOATING ARGUMENTS () () *) +) *) +) ') ') ST(AF) = { {', +}, {(, +} } ! + is justified ODD-LENGTH CYCLES: A PROBLEM FOR STABLE SEMANTICS ') No stable extension exists! () *) (and also imposing ø is not satisfactory) 37
  • 38. Stable Semantics: an unsatisfactory patch Stable extensions = - conflict-free sets attacking all outside arguments, if there is one - {ø}, otherwise '1) '3) () *) '2) ST(AF) = {ø } ! ( NOT justified!!! 38
  • 39. Preferred semantics Stable extensions are maximal complete extensions •  conflict-free: by definition •  admissible: every argument attacking an extension is outside ! attacked by the extension itself •  maximal: no argument can be included! Preferred semantics [P.M. Dung, 95] Preferred extension Maximal complete extension = max Set: •  is conflict-free •  defends all of its elements 39
  • 40. Preferred semantics and floating arguments () *) +) () () *) +) ') *) +) ') () ') *) +) ') PR(AF) = ST(AF) = { {', +}, {(, +} } ! + is justified () Grounded semantics: *) +) ') 40
  • 41. Preferred semantics and odd-length cycles ') PR(AF) = {ø} () A big difference, isn’t it? ST(AF) = ø *) GE(AF) = {ø} No argument justified w.r.t. grounded and preferred semantics •  As stable semantics, preferred semantics handles the case of floating arguments (differently wrt grounded semantics) •  W.r.t. stable semantics it behaves “better” in the case of odd-length cycles (as the grounded semantics) So, what remains to be done? 41
  • 42. Semi-stable semantics (1) •  Stable semantics - clashes in some cases (odd-length cycles), however: - a widely applied approach (default logic, stable models of logic programming, answer set programming, etc.) - a very credulous approach: stable extensions are preferred but not viceversa ! justified arguments w.r.t. stable semantics are a (sometimes strict) superset of arguments justified w.r.t. preferred semantics, e.g. ,) ,) *) ') () *) ') () +) +) PR(AF)={{', !}, {"}} ST(AF) = {{', !}} 42
  • 43. Semi-stable semantics (2) •  Aims at guaranteeing existence of extensions [Verheij’96, (differently from stable semantics) Caminada’06] + coinciding with stable semantics when stable extensions exist (differently from preferred semantics) •  Definition: E- SST(AF) iff E is a complete extension such that (E U {'| E% '}) is maximal •  Main properties: -  A semistable extension always exists (in the finite case!) since a maximization requirement replaces “aggressive attack” -  If a stable extension E exists, then (E U {'| E% '}) includes all arguments, therefore semistable extensions # stable extensions -  In any case, semistable extensions are preferred extensions, but the opposite is not always true 43
  • 44. Semi-stable semantics: examples Example for existence ') The unique admissible set is empty () ! trivially maximizes (E U {'| E% '} ) *) Example for backward compatibility (and difference w.r.t. preferred semantics) ,) PR(AF)={{', !}, {"}} *) ') () SST(AF)={{', !}} )= ST(AF) +) 44
  • 45. CF2 semantics: motivation Preferred/stable/semistable semantics and cycles ') ') () () ') () *) ') ') () () *) A different treatment for even and odd-length cycles. Is it just a matter of symmetry and elegance? 45
  • 46. Preferred/Semistable Semantics and cycles PR(AF) = ') () +1) +2) {{', +1}, {', +2}, VS {(, +2} } 46
  • 47. Preferred/Semistable Semantics and cycles PR(AF) = ') () +1) +2) {{', +1}, {', +2}, VS {(, +2} } () *) +1) +2) PR(AF) = {{+2}} ') VS 47
  • 48. Preferred/Semistable Semantics and cycles PR(AF) = ') () +1) +2) {{', +1}, {', +2}, VS {(, +2} } () *) +1) +2) PR(AF) = {{+2}} ') VS () PR(AF) = {{', *, +2}, ') *) +1) +2) {(, +, +1}, {(, +, +2} } +) NOTE: grounded semantics yields the empty set in all cases 48
  • 49. Pollock example revisited (1) Jones Smith It’s It’s not unreliable unreliable raining raining Rob unreliable Rob says Smith says Jones says Smith says Bob says Jones unrel. Rob unrel. Smith unrel. it’s raining it’s not raining 49
  • 50. Pollock example revisited (2) Fred says Jones unrel. Jones unreliable Fred Smith It’s It’s not unreliable unreliable raining raining Rob unreliable Rob says Smith says Jones says Smith says Bob says Fred unrel. Rob unrel. Smith unrel. it’s raining it’s not raining 50
  • 51. Preferred Semantics and Floating Arguments again… () [ two preferred *) +) extensions] ') VS () [empty set is the unique *) +) .) preferred extension] ') NB: same behavior for semistable semantics, stable semantics clashes, grounded semantics yields the empty set in both cases 51
  • 52. Strongly connected components (SCCs) Equivalence classes under the relation of path-equivalence (mutual reachability) () *) .1) .2) ') () *) .1) .2) ') () *) .1) .2) ') 52
  • 53. Strongly connected components (SCCs) SCCs form an acyclic graph S3 S6 S1 S4 S2 S5 S7 S1 and S2 are initial SCCs S1 is sccparent of S3, S4 and S5 all other SCCs precede S7 53
  • 54. CF2 semantics: the definition E- CF2(AF) iff: - E - MCF(AF) if |SCCSAF| = 1 - / S - SCCSAF (E0S) - CF2(AF UP_AF(S,E)) otherwise S UP_AF(S,E) 54
  • 55. CF2 semantics and odd-length cycles (1) () *) ') () () () *) *) *) ') ') ') Maximal conflict-free sets 55
  • 56. CF2 semantics and odd-length cycles (2) () *) .1) .2) ') {*,.2}, {',.1}, {',.2}, {(,.1}, {(,.2} Yields several extensions ! all arguments not justified in both cases () ') *) .1) .2) +) {',*,.2}, {(,+,.1}, {(,+,.2} 56
  • 57. Floating arguments with a three-length cycle () *) +) .) ') () () *) +) .) *) +) .) ') ') () *) +) .) ') Extensions: {*,.}, {',.}, {(,.} Defeat status 57
  • 58. since we do not want problems in one in relation toknowledge base to affect other, advantageous part of the consistency requirements, as explained in the following. ly unrelated parts of the knowledge base. generates an argumentation framework based on a set of propositional formul Suppose one Stable semantics is therefore not an option. semantics have to be admissibility based? That is,rules desirable that each the propositional formulas express informatio P and a set of defeasible is it D. The idea is that extension an admissible (or even complete) one? Again, it ofthe defeasible rules an ultimate of thumb that can be subject that isA problem is difficult to provide express(1) beyond doubt and CF2 semantics rules neral: one has to refer to specific contexts. In particular, following knowledge base: exceptions. Now consider the in the context of instantiated nerated from an underlying logical knowledge base, admissibility can be regarded as P ¼ fjw; mw; sw; :ðjt ^ mt ^ stÞg s in relation to consistency requirements, as explained in the following. ne generates•  Considering some examples with structuredmw ) mt; sw ) it turns out that an argumentation framework based on a set ¼ fjw ) jt; arguments, stg D of propositional formulas f defeasible rules D. The idea is that the propositional formulas express information and Suzy want to go cycling on conflict-freenessexpress not entail consistency, e.g. Mary, This example can be interpreted as follows: John, nd doubt and the defeasible rules doesthat John wants to thaton thebe subject tois a reason to believe that John will b tandem. The fact rules of thumb get can tandem (jw) An introduction to argumentation semantics Now consider the following knowledge base: The same holds for Mary and Suzy. However, since the tandem only has tw 407 on the tandem (jt). seats, they :ðjt ^ mt ^ it with the three of them: :(jt 4 mt 4 st). From this knowledge base, w P ¼ fjw; mw; sw;cannot be onstÞg A5 can then construct the 10 following arguments, based on an argument construction scheme D ¼ fjw ) jt; mw ) mt; sw ) stg Amgoud (2007) and Prakken (2010): presented in Caminada and A1 5 :(jt 4 mt 4 st) ple can be interpreted as follows: John, Mary, and Suzy want to go cycling on a A2 A8 A3 A 5 tandem (jw) is a reason to believe that John will be fact that John wants to get on2thejw A3 5 mw m (jt). The same holds for Mary and Suzy. However, since the tandem only has two nnot be on it with the three A4 5 sw :(jt 4 mt 4 st). From this knowledge base, we of them: A1 A5 5 A2 ) jt on an argument construction scheme as struct the 10 following arguments, based A10 A9 Caminada and Amgoud (2007)5 A3 ) mt A6 and Prakken (2010): A7 5 A4 ) st 4 mt 4 st) A8 5 A6, A7, A1 - :jt A9 5 A5, A7, A1 - :mt A4 A7 A6 A10 5 A5, A6, A1 - :st ) jt Assuming the principle of restrictednot enough to obtain consistent conclusions Figure 20 Conflict-freeness is rebutting23 it would then follow that A8 attacks A5, A9, an A10, that A9 attacks A6, A8, and A10, and that A10 attacks A7, A8, and A9. This yields th 58 mt argumentation also semi-stable and preferred extensions). It should be mentioned that the sets of conclusions framework of Figure 20.
  • 59. A problem of CF2 semantics (2) •  By slightly complicating example, one can find a CF2 extension which is not consistent •  On the other hand, admissibility entails consistency (proved by Caminada & Amgoud 2007, AIJ) ADMISSIBILITY CAN BE A DESIRED REQUIREMENT 59
  • 60. TO CONCLUDE… GROUNDED PRUDENT STABLE STAGE PREFERRED ROBUST SEMISTABLE TOLERANT CF2 SUSTAINABLE IDEAL EACH SEMANTICS HAS ITS OWN ROLE… … WHICH ONE IS A GOOD RESEARCH QUESTION…! 60
  • 61. What about general principles? Here we consider only some semantics - see [Baroni & Giacomin ’06] Grounded Preferred CF2 Semistable CF-principle Yes Yes Yes Yes Admissibility Yes Yes No Yes Reinstatement Yes Yes No Yes Weak reinstatement Yes Yes Yes Yes CF-reinstatement Yes Yes Yes Yes I-maximality Yes Yes Yes Yes Directionality Yes Yes Yes No Weak Skepticism Yes No Yes No Adequacy [all forms] Weak Resolution No Yes No Yes Adequacy [all forms] 61
  • 62. Applications and principles PRINCIPLES TO BE TO BE STUDIED DEEPENED APPLICATION DOMAINS SEMANTICS 62
  • 63. Semantics and attitude SKEPTICAL REASONING CREDULOUS REASONING E E E1 W E2 : E1 C E2 : / E2 -E2, 1 E1 -E1 : E1" E2 / E1 -E1, 1 E2 -E2 : E1" E2 63 Fig. 3. S +, S → and S relations for any argumentatio
  • 64. MANY THANKS FOR YOUR KIND ATTENTION 64
  • 65. Selected references (1) Landmark argumentation papers and books S. Toulmin, “The Uses of Argument” Cambridge University Press, 1958. R. P. Loui, “Defeat Among Arguments: a System of Defeasible Inference”, Computational Intelligence, vol. 3(3), 1987. J. Pollock, “Defeasible Reasoning”, Cognitive Science, vol. 11(4), 1987. G. Simari & R. P. Loui, “A mathematical treatment of defeasible reasoning and its implementation , Artificial Intelligence, vol. 53(2-3), 1992. Argumentation surveys H. Prakken & G.A.W. Vreeswijk, “Logics for Defeasible Argumentation”, in Handbook of Philosophical Logic, 2nd Edition, Kluwer Academic Publishers, 2001. C.I. Chesnevar, A.G. Maguitman, R.P. Loui, “Logical models of argument”, ACM Computing Surveys, vol. 32(4), 2000. 65
  • 66. Selected references (2) Argumentation semantics survey P. Baroni, M. Caminada, M. Giacomin “An introduction to argumentation semantics , The Knowledge Engineering Review, vol. 26(4),2011. Books D. Walton, “Fundamentals of critical argumentation , Cambridge University Press, 2006. P. Besnard & A. Hunter, “Elements of Argumentation , MIT Press, 2008. “Argumentation in Artificial Intelligence , edited by I. Rahwan and G. R. Simari, Springer, 2009. Dung s influential paper on abstract argumentation P.M. Dung, “On the Acceptability of Arguments and Its Fundamental Role in Nonmonotonic Reasoning, Logic Programming, and n-Person Games , Artificial Intelligence, vol. 77(2), 1995. 66
  • 67. Selected references (3) Semantics P. Baroni, M. Giacomin, G. Guida, “SCC-recursiveness: a general schema for argumentation semantics” Artificial Intelligence, vol. 168(1-2), 2005. B. Verheij, “Two approaches to dialectical argumentation:admissible sets and argumentation stages”, Proc. of the 8th Dutch Conference on Artificial Intelligence, 1996 M. Caminada, “Semi-Stable Semantics”, Proc. of 1st International Conference on Computational Models of Arguments (COMMA 2006), 2006 P.M. Dung, P. Mancarella, F. Toni, “A dialectic procedure for sceptical, assumption-based argumentation”, Proc. of 1st International Conference on Computational Models of Arguments (COMMA 2006), 2006 S. Coste-Marquis, C. Devred, P. Marquis, "Prudent Semantics for Argumentation Frameworks", Proc. of 17th IEEE International Conference on Tools with Artificial Intelligence (ICTAI 2005), 2005 67
  • 68. Selected references (4) Semantics H. Jakobovits & D. Vermeir, "Robust Semantics for Argumentation Frameworks", Journal of Logic and Computation 9(2), 1999 P. Baroni, M. Giacomin, G. Guida, “SCC-recursiveness: a general schema for argumentation semantics” Artificial Intelligence, vol. 168(1-2), 2005. P. Baroni, M. Giacomin, “Resolution-based argumentation semantics”, Proc. of 2nd International Conference on Computational Models of Arguments (COMMA 2008), 2008 G.A. Bodanza, F.A. Tohmé, “Two approaches to the problems of self-attacking arguments and general odd-length cycles of attack” Journal of Applied Logic, to appear. P. Baroni, P. Dunne, M. Giacomin, “Computational Properties of Resolution-based Grounded Semantics”, IJCAI 2009, to appear. 68
  • 69. Selected references (5) General criteria for semantics evaluation and comparison M. Caminada & L. Amgoud, “On the evaluation of argumentation formalisms”, Artificial Intelligence, vol. 171(5-6), 2007. P. Baroni, M. Giacomin, G. Guida, “On principle-based evaluation of extension-based argumentation semantics”, Artificial Intelligence, vol. 171(10-15), 2007. P. Baroni, M. Giacomin, Skepticism relations for comparing argumentation semantics , International Journal of Approximate Reasoning, vol. 50(6), 2009. 69
  • 70. !"#$%&'()*+',-+",!#&'(,.+/0&*&1, 2)"(,33, 4%33(5()(%$/&6(%7/5($& !"#"$%&'())%*%& 8$(9"#3(*%:&;(&<#"37(%& +,-+.&!/01(%&.$20/)(3&
  • 71. !"#$%&'()*+',-+",:&/010+',;)<0'#, !"#$%&'()*+',-+",4&50&-1,6&7010+', !"#$%&'()*+',-+",6&)1+'0'#,, )8+$(,9"$1(,
  • 73. !"#$%&'()*+',-+",=&/010+',%)<0'#, •  :&/010+',%)<0'#>, –  .&5&/(,(?&,8&1(,+",#++=,)/*+'1@, –  !/*+'1,-&)1085&,)%+'#,)5(&"')*7&1A, •  B&,'&&=,(+,<'+C, –  3'-+"%)*+',D0-,)7)05)85&E,)8+$(,(?&,/$""&'(,1()(&, +-,(?&,C+"5=@, –  F+'1&G$&'/&1,+-,H+(&'*)5,)/*+'1A,,
  • 74. !"#$%&'()*+',-+",=&/010+',%)<0'#, •  2"+85&%1>, –  !7)05)85&,0'-+"%)*+',0'/+%H5&(&, –  !7)05)85&,0'-+"%)*+',H&"7)=&=,C0(?,$'/&"()0'(I, =/>&5?71&*1"&7/$3"@?"$7"3&/A&%$&%72/$&B*&*1"& 0#"A"#"$7"3&/A&*1"&;"7(3(/$&5%C"#D&
  • 75. !"#$%&'()*+',-+",=&/010+',%)<0'#, •  F5)110/)5,=&/010+',(?&+"I, –  2"0'/0H5&1,-+",/+%H)"0'#,=0J&"&'(,)5(&"')*7&1A, –  3'H$(1>, •  .&(,+-,/)'=0=)(&,)/*+'1@, •  K$'/*+',(+,)11&11,(?&,7)5$&,+-,(?&0",/+'1&G$&'/&1@, •  F+%H5&(&,+",H)"*)5,0'-+"%)*+',)8+$(,1()(&,+-,C+"5=A, –  L$(H$(>,H"&-&"&'/&,"&5)*+',8&(C&&',)/*+'1A,
  • 76. !"#$%&'()*+',-+",=&/010+',%)<0'#, •  M'+C5&=#&,)'=,2"&-&"&'/&1>, –  :01("08$*+',-$'/*+',)11&110'#,H5)$108050(I,+-,(?&, 1()(&1,+-,(?&,C+"5=@, –  N*50(I,-$'/*+',&'/+=0'#,H"&-&"&'/&1,(+,&1*%)(&, (?&,#++='&11,+-,),/+'1&G$&'/&A, •  2"0'/0H5&1,$1&=,(+,/+%H)"&,H)0"1,+-, )5(&"')*7&1A,
  • 77. !"#$%&'()*+',-+",=&/010+',%)<0'#, •  !')5I*/)5,&OH"&110+'1,1$%%)"0P&,(?&, =&/010+'A, •  Q)"=,(+,$'=&"1()'=,C?I,),H"+H+1&=, )5(&"')*7&,01,#++=@,+",8&R&"A, •  B&,'&&=,)',)HH"+)/?,(+,$'=&"1()'=,(?&, $'=&"H0''0'#,+-,(?&,&7)5$)*+'A,
  • 78. !"#$%&'()*+',-+",=&/010+',%)<0'#, •  B?I,$10'#,)"#$%&'()*+'S, –  TOH5)')(+"I,H+C&"@, –  :&/010+'1,8)1&=,+',)"#$%&'(1,)'=,,,,,,,,,,,,,, /+$'(&"U)"#$%&'(1@, –  !85&,(+,&OH5)0',!"#,),/?+0/&,?)1,8&&',%)=&A,
  • 79. 9C+,)5(&"')*7&,)HH"+)/?&1, •  !%#+$=@,2")=&A,83($E&%#E?5"$*3&A/#&5%C($E& %$;&"F0)%($($E&;"7(3(/$3G,!"*-A,3'(&55A,VWXDXUYE>, YVXUYXZ,D[]EA, •  M)<)1@,;+")0*1A,.#E?5"$*%2/$&H%3";&;"7(3(/$& 5%C($E&A/#&%?*/$/5/?3&%E"$*3G&3',2"+/1,+-, !!;!.,[X>,^^XU^]A,
  • 80. !%#+$=,_,2")=&,)HH"+)/?, •  :&`'&,),D/+%H5&(&E,H"&U+"=&"0'#,+',),1&(,+-, H+11085&,+H*+'1,:A, •  !"#$%&'()*+'U8)1&=,=&/010+',%)<0'#>, –  F+'1("$/*'#,)"#$%&'(1,0',-)7+"a)#)0'1(,1()(&%&'(1, DH&"()0'0'#,(+,8&50&-1,+",=&/010+'1E@, –  T7)5$)*'#,(?&,1("&'#(?,+-,&)/?,)"#$%&'(@, –  :&(&"%0'0'#,(?&,=0J&"&'(,/+'b0/(1,)%+'#,)"#$%&'(1@, –  T7)5$)*'#,(?&,)//&H()8050(I,+-,)"#$%&'(1@,, –  F+%H)"0'#,=&/010+'1,+',(?&,8)101,+-,"&5&7)'(, c)//&H(&=d,)"#$%&'(1A,
  • 81. $%&'()*+ •  ,*-./.012,?)70'#,),1$"#&"I,D1#E,+",'+(,De1#EA, •  310!)*45*2,(?&,H)*&'(,?)1,/+5+'0/,H+5IH1A, •  9?&,M4,/+'()0'1,(?&,-+55+C0'#,0'-+"%)*+'>, –  ?)70'#,),1$"#&"I,?)1,10=&U&J&/(1f, –  '+(,?)70'#,1$"#&"I,)7+0=1,?)70'#,10=&U&J&/(1f,, –  C?&',?)70'#,),/)'/&"@,?)70'#,),1$"#&"I,)7+0=1,5+11,+-,50-&f, –  0-,),H)*&'(,?)1,/)'/&",)'=,?)1,'+,1$"#&"I@,(?&,H)*&'(,C+$5=,5+1&,?01, 50-&f, –  (?&,H)*&'(,?)1,/+5+'0/,H+5IH1f, –  ?)70'#,/+5+'0/,H+5IH1,%)I,5&)=,(+,/)'/&"f, g+)51,50<&>,c'+,10=&,&J&/(1d,)'=,c(+,'+(,5+1&,?01,50-&dA,, ;+"&,0%H+"()'(,-+",?0%,(+,'+(,5+1&,?01,50-&,(?)',(+,'+(,?)7&,10=&,&J&/(1A,
  • 82. 9?&,-")%&C+"<, •  !81(")/(,)"#$%&'(U8)1&=,-")%&C+"<A, •  9C+,1(&H1>, –  .(&H,V, •  !"#$%&'(1,-+",8&50&-1@, •  !"#$%&'(1,-+",+H*+'1@, •  4$05=0'#,)'=,&7)5$)*+',$'=&",/5)110/)5,1&%)'*/1A, –  .(&H,[, •  F+%H)"01+',+-,H)0"1,+-,+H*+'1,$10'#,=&/010+', H"0'/0H5&1A,
  • 83. 9?&,-")%&C+"<, •  9?"&&,/5)11&1,+-,=&/010+',H"0'/0H5&1>, –  N'0H+5)">,, •  L'5I,)"#$%&'(1,H"+1@, •  L'5I,)"#$%&'(1,/+'1A, –  40H+5)", •  4+(?,(IH&1A, –  h+'UH+5)", •  !##"&#)*+',+-,H"+1,)'=,/+'1,)"#$%&'(1,0'(+,%&()U )"#$%&'(A,
  • 84. 9?&,-")%&C+"<, •  :&/010+'1,1$HH+"(&=,8I,)"#$%&'(1>, –  &%H?)10P0'#,0(1,H+10*7&,/+'1&G$&'/&1@,, –  #+)51,1)*1`&=@, –  "&i&/*+'1,)7+0=&=A, •  :&/010+'1,)R)/<&=,8I,)"#$%&'(1>,, –  &%H?)10P0'#,0(1,'&#)*7&,/+'1&G$&'/&1@,, –  %011&=,#+)51@, –  /&"()0',"&i&/*+'1A,
  • 85. 9C+,)5(&"')*7&,)HH"+)/?&1, •  !%#+$=@,2")=&A,83($E&%#E?5"$*3&A/#&5%C($E& %$;&"F0)%($($E&;"7(3(/$3G&!"*-A,3'(&55A,VWXDXUYE>, YVXUYXZ,D[]EA, •  M)<)1@,;+")0*1A,.#E?5"$*%2/$&H%3";&;"7(3(/$& 5%C($E&A/#&%?*/$/5/?3&%E"$*3G&3',2"+/1,+-, !!;!.,[X>,^^XU^]A,
  • 86. M)<)1,_,;+")0*1,)HH"+)/?, •  ;+=$5)",)#&'(,)"/?0(&/($"&>, –  ;+=$5&1,=&=0/)(&=,(+,)#&'(j1,/)H)8050*&1@, –  ;+=$5&1,)"&,0'=&H&'=&'(A, •  L7&")55,8&?)70+$",+-,(?&,)#&'(>, –  3'(&")/*+'1,)%+'#,=0J&"&'(,%+=$5&1@, –  3'(&")/*+'1,)%+'#,(?&0",1&H)")(&,=&/010+'1@, –  :&508&")*+',H"+/&11,C0(?0',&)/?,%+=$5&A,
  • 87. 9?&,%+=&5, •  !"#$%&'()*7&,=&508&")*+',%+=$5&A, •  k)"0+$1,=&/010+',%)<0'#,("&)(&=,$'0-+"%5IA, •  Q0#?,5&7&5,+-,)=)H()8050(I,C?&',(?&, &'70"+'%&'(,/?)'#&1A, •  T'/+%H)110'#,(?&,0'b$&'/&,+-, –  ,=0J&"&'(,"&5)*7&,"+5&1,+-,0'(&")/*'#,)#&'(1@,)'=, –  ,(?&,/+'(&O(,+-,(?&,H)"*/$5)",0'(&")/*+'A,
  • 88. 9?&,%+=&5, •  6+5&1,)'=,/+'(&O(,)1,=I')%0/,H"&-&"&'/&1,+', (?&,=&/010+',H+50/0&1A, •  6+5&1,)'=,/+'(&O(,"&H"&1&'(&=,0',(C+,%+=$5)", H)"(1A, •  !0%0'#,(+,H"+70=&,(?&,)#&'(,C0(?,"+8$1('&11>, –  0'/+%H5&(&,0'-+"%)*+',-"+%,&'70"+'%&'(,!, )8=$/*+',C0(?0',)"#$%&'()*+',-")%&C+"<f, –  )#&'(,)85&,(+,=&508&")(&,+',)5(&"')*7&,/?+0/&f, –  )#&'(,()<&,=&/010+'1,/+'=0*+')5,+',)11$%H*+'1, )8+$(,&'70"+'%&'(f,
  • 89. 9?&,%+=&5, •  2&"1+')50(I,%+=$5&, –  8)1&=,+',(?&,1)%&,)"#$%&'()*+',-")%&C+"<, $1&=,-+",(?&,+(?&",%+=$5&1@, –  "&H"&1&'*'#,(?&,H&"1+')50(I,+-,)',)#&'(,)1,), =&/010+',H+50/I,)//+"=0'#,(+,'&&=1,)'=, %+*7)*+'1,+-,)#&'(1@, –  )==0*+')5,i$=#%&'(,+',=&/010+',H"+85&%1,+-,)'I, +'&,+-,(?&,+(?&",%+=$5&1,=&H&'=0'#,+',7)"0+$1, '&&=1A,
  • 90. 9?&,%+=&5, •  9?"&&,5&7&51,0',)',)#&'(j1,(?&+"I>, –  K0"1(,5&7&5>,"$5&1,9,(?)(,"&-&",=0"&/(5I,(+,(?&,1$8i&/(, =+%)0',+-,(?&,)#&'(,D678*-9:)*;*)+,*-./.01+<=)*/+ 0>+9"*+&5*19?+ –  .&/+'=,)'=,(?0"=,5&7&5>,"$5&1,"&5)(&,(+,(?&,H+50/I, $'=&",C?0/?,(?&,)#&'(,$1&1,?01,+8i&/(U5&7&5, =&/010+',"$5&1,)//+"=0'#,(+,"+5&1,@AB+C+<0)*+ AB.0B.D*/?E+)'=,/+'(&O(,@A-+:+F019*%9+AB.0B.D*/?G,
  • 91. $%&'()*+ "V,>,"&G$&1(l?&5HD!@9@!VE,m,'&&=l/++H&")*+'D!@9E@, "&5&7)'(D9@!VE, "[,>,e"&G$&1(l?&5HD!@9@,!VE,m,"&G$&1(l?&5HD!@9@,![E@,, ,!V,n,![,, 6V,>,?lHD"VD!@9@!VE@"VD!@9@![EE,m,%)')#&%&'(l()1< D9E@%)')#&"D!VE,, 6[,>,?lHD"VD!@9@,![EE@,"VD!@9@,!VEE,m,(&/?'0/)5l()1< D9E@&OH&"(D![E,, FV,>,?lHD6V@6[E,m,%)"<&(l1?)"&l0'/"&)1&lH&"0+=, F[,>,?lHD6[@6VE,m,e%)"<&(l1?)"&l0'/"&)1&lH&"0+=,
  • 93. !"#$%&'()*+',-+",8&50&-,"&7010+', •  H*).*>+B*;./.01+01,(?&,H"+/&11,+-,/?)'#0'#,8&50&-1, (+,)=)H(,(?&,&H01(&%0/,1()(&,+-,)',)#&'(,(+,),'&C, H0&/&,+-,0'-+"%)*+'A, •  IB5='*19&D01,01,/+'/&"'&=,C0(?,(?&,&7)5$)*+', +-,/5)0%1,8)1&=,+',H"&%01&1,0',+"=&",(+,"&)/?, /+'/5$10+'1A, •  H09"+0>+9"*'>,, –  H"+70=&,8)10/,)'=,1$81()'*)5,(&/?'0G$&1,-+",(?&,)"(,+-, "&)1+'0'#@,)1,H&"-+"%&=,8I,?$%)',8&0'#1,0',&7&"I=)I, 50-&,10($)*+'1@, –  #+,-)",8&I+'=,5+#0/)5,=&=$/*+'A,
  • 94. !"#$%&'()*+',-+",8&50&-,"&7010+', •  !#&'(,&'70"+'%&'(@, •  4&50&-,"&7010+',=&1/"08&1,, –  (?&,C)I,0',C?0/?,)',)#&'(,01,1$HH+1&=,(+,/?)'#&,?&", 8&50&-1,C?&','&C,0'-+"%)*+',)""07&1@,+",/?)'#&1,0', (?&,C+"5=,)"&,+81&"7&=f,, •  !"#$%&'()*+',=&)51,, –  C0(?,1(")(&#0&1,)#&'(1,&%H5+I,-+",(?&0",+C',"&)1+'0'#@,, –  (+,/?)'#&,(?&,8&50&-1,+-,+(?&",)#&'(1@,8I,H"+70=0'#, "&)1+'1,-+",1$/?,/?)'#&A,
  • 95. !"#$%&'()*+',-+",8&50&-,"&7010+', !"#$%&'()*+',)'=,8&50&-,"&7010+',)1, /+%H5&%&'()"I,=01/0H50'&1A, T)/?,'&&=1,(?&,+(?&"j1,1$HH+"(,(+,%+=&5, 1$//&11-$5,=&/010+',%)<0'#,0',"&)5,C+"5=, )HH50/)*+'1A,
  • 96. !"#$%&'()*+',_,4&50&-,6&7010+', •  !',)"#$%&'(,!,-+",o,01,),1&(,+-,0'(&""&5)(&=, H0&/&1,+-,<'+C5&=#&,1$HH+"*'#,o,-"+%, &70=&'/&A, •  F5)110/)5,8&50&-,"&7010+'>,`O&=,`'0(&,5)'#$)#&,p, C0(?,),/+%H5&(&,1&(,+-,8++5&)',/+''&/*7&1A, •  ;)'I,=0J&"&'(,-")%&C+"<1,-+",8&50&-,"&7010+', C0(?,(?&0","&1H&/*7&,&H01(&%0/,%+=&51A,,
  • 97. 4&50&-,6&7010+', •  $(./9*'.-+'04*)>,-+"%)501%,0',C?0/?,8&50&-1,)"&, "&H"&1&'(&=@,)'=,0',C?0/?,=0J&"&'(,<0'=1,+-, +H&")(+"1,/)',8&,=&`'&=A,, •  4)10/,"&H"&1&'()*+',+-,&H01(&%0/,1()(&1>, –  7*).*>+/*9/>,1&(1,+-,1&'(&'/&1,/5+1&=,$'=&",5+#0/)5, /+'1&G$&'/&@,+",, –  7*).*>+7&/*/>,1&(1,+-,1&'(&'/&1,'+(,'&/&11)"05I,/5+1&=A, –  LH&")(+"1,H"&1&'(&=,0',(C+,C)I1>,, •  8I,#070'#,)',&OH50/0(,/+'1("$/*+',D)5#+"0(?%E,-+",(?&, +H&")(+"@,+", •  8I,#070'#,),1&(,+-,")*+')50(I,H+1($5)(&1,(+,8&,1)*1`&=, D/+'1(")0'(1EA,,
  • 98. !"#$%&'()*+',_,4&50&-,6&7010+', •  <*-*.;.15+1*!+.1>0B'&D01>, –  '&C,0'-+"%)*+',0',=0J&"&'(,1?)H&1,)'=,-+"%1@,&A#A@,3,01,),H"+H+10*+')5,-)/(f, –  8)10/,!g;,(?&+"I@,)11$%0'#,(?&,&H01(&%0/,1()(&,+-,(?&,)#&'(,(+,8&,#07&',8I,),8&50&-,1&(f, –  3,%0#?(,8&,%+"&,/+%H5&O@,&G$0HH&=,C0(?,=&#"&&,+-,H5)$108050(I@,+",?)7&,-+"%,+-,),"$5&@,+",), /+%H5&(&,)"#$%&'(@,+",+-,),1&(,+-,1$/?,&'**&1f, •  $;&)=&D15+1*!+.1>0B'&D01>,, –  -+",-$"(?&",H"+/&110'#,3@,/"$/0)5,-+",(?&,)#&'(,(+,<'+C,0(1,+"0#0'@,)1,(?01,<'+C5&=#&,0'b$&'/&1, C0550'#'&11,(+,)=+H(,3f, –  3,8)1&=,+',)',+81&"7)*+',%)=&,8I,)#&'(,?&"1&5-@,1?&,C055,$1$)55I,8&,/+'70'/&=,+-,0(,8&0'#,("$&f,, –  0-,3,01,/+'7&I&=,8I,)'+(?&",)#&'(@,(?&,)#&'(,C055,"&G$0"&,1+%&,i$1*`/)*+',-+",3f, –  )1,%)'=)(+"I,1(&H,-+",")*+')5,(?0'<0'#@,1?&,&7)5$)(&,8+(?,3,)'=,H+11085&,i$1*`/)*+',+',(?&, 8)101,+-,?&",+C',8&50&-1@,)'=,=&/0=&,0-,3,01,(+,8&,0'/+"H+")(&=,0'(+,?&",8&50&-1,+",'+(f, •  +F"&15.15+7*).*>/>,, –  0-,(?&,)#&'(,=&/0=&=,(+,)=+H(,3@,1?&,&%H5+I1,1(")(&#0&1,(+,0'/+"H+")(&,3,/+'101(&'(5I,0'(+,?&", 8&50&-1A,K+",(?01@,1?&,?)1,(+,$1&,8&50&-,"&7010+',(&/?'0G$&1,(+,/?)'#&,?&",&H01(&%0/,1()(&f, •  J1>*B*1-*>,, –  K"+%,'&C,&H01(&%0/,1()(&@,(?&,)#&'(,=&"07&1,H5)$1085&,8&50&-1,(?)(,#$0=&,?&",8&?)70+$"f,
  • 99. !"#$%&'()*+',_,4&50&-,6&7010+', •  T%8&==0'#,0'(+,/+%H5&O,"&)1+'0'#,H"+/&11A, •  9?&,/+%H5&%&'()"I,/?)")/(&"1,+-,)"#$%&'()*+', )'=,8&50&-,"&7010+',8&/+%&,&70=&'(>, –  )"#$%&'()*+',%)<&1,/+'("08$*+'1,(+,(?&,&7)5$)*+', 1(&H@, –  8&50&-,"&7010+',01,&%H5+I&=,0',(?&,8&50&-,/?)'#&,H)"(A,, –  T7)5$)*+',0'/5$=0'#,?IH+(?&*/)5,/?)'#&, H"+/&11&1>, •  C?)(,C+$5=,?)HH&',0-,(?&,'&C,0'-+"%)*+',C&"&,(+,8&, 8&50&7&=@,, •  8&50&-,/?)'#&,0%H50/0(5I,"&50&1,+',5+#0/)5,50'<1,8&(C&&',H0&/&1, +-,0'-+"%)*+',C?0/?,/)',8&,"&H"&1&'(&=,8I,)"#$%&'(1A,,
  • 100. !"#$%&'()*+',_,4&50&-,6&7010+', •  4+(?,-"+%,)"#$%&'()*+',H"+/&11&1,)'=,-"+%, 8&50&-,"&7010+',H"+/&11&1@,H5)$1085&,8&50&-1,/)', 8&,+8()0'&=A, •  4+(?,)"&)1,-+/$1,+'5I,+',H)"(1,+-,(?&,=I')%0/, "&)1+'0'#,H"+/&11,C?05&,)(,(?&,1)%&,*%&, H"+70=0'#,#&'&")5,)'=,7&"1)*5&,-")%&C+"<1A,,
  • 101. F+%H)"01+', •  ,.K*B*1-*/+7*9!**1+&B5='*19&D01+&14+7*).*>+B*;./.01>, –  "&H"&1&'()*+')5,011$&1@,1I'()/*/,)'=,1&%)'*/,-+$'=)*+'1,+-,8+(?,)"&)1f,, –  0',1()'=)"=,8&50&-,"&7010+'@,5+#0/)5,-+"%$5)1,$1&=,-+",M6@,"&1$5(1,+-,/?)'#&, H"+/&11&1,)"&,5+#0/)5,-+"%$5)1f, –  &H01(&%0/,1()(&1,)"&,/?)'#&=,C0(?,"&1H&/(,(+,&'("&'/?%&'(@,H5)$108050(I@,qf, –  (+,7&"0-I,"&1$5(1,+-,/?)'#&,H"+/&11&1@,/5)110/)5,5+#0/)5,1&%)'*/1,01,$1&=f, –  )"#$%&'()*+',-+/$1&1,+',0'(&")/*+'1,+-,)"#$%&'(1,)1,H0&/&1,+-,0'-+"%)*+', (?)(,%)I,)R)/<,+'&,)'+(?&"@,)'=,"&5)*+',8&(C&&',)"#$%&'(1,%)I,#07&, H"0+"0(I,(+,+'&,)"#$%&'(,+",)'+(?&"f, –  )"#$%&'(1,(?&%1&57&1,)"&,7&"I,?&(&"+#&'&+$1f, –  1&%)'*/1,%)<&1,H"&/01&,C?)(,#++=,)"#$%&'(1,)"&f, •  F0''01+5B0=14/>, –  8+(?,=01/0H50'&1,)0%,)(,"&1+570'#,/+'b0/(1,C?0/?,)"&,$1$)55I,8)1&=,+',5+#0/)5, #"+$'=1@,0A&A@,+',/+'(")=0/*+'1f, –  %)<&,$1&,+-,H"&-&"&'/&,"&5)*+'1,(+,)/?0&7&,(?01,)0%f, –  8&50&-,"&7010+',H"+70=&1,),?0#?5I,=&/5)")*7&,-")%&C+"<,-+",(?)(@,8)1&=,+', H+1($5)(&1f, –  )"#$%&'()*+',01,%+"&,/+'/&"'&=,C0(?,H")/*/)5@,i$1*`/)*+'U8)1&=, (&/?'0G$&1f,
  • 102. !"#$%&'()*+',0',4&50&-,6&7010+', •  Q+C,)"#$%&'()*+',(&/?'0G$&1,/)',8&,$1&=,0', 8&50&-,"&7010+',(?&+"IS, –  r$1*`/)*+'U8)1&=,("$(?,%)0'(&')'/&,1I1(&%1,s:+I5&t>, 0'(&")/*+'1,8&(C&&',i$1*`/)*+'1,C?&',),'&C, i$1*`/)*+',?)1,8&&',)==&=@,(+,`'=,+$(,C?0/?,/+'/5$10+'1, /)',8&,i$1*`&=A, –  !11$%H*+'U8)1&=,("$(?,%)0'(&')'/&,1I1(&%1,sK)5)HH),&(, )5At>,%)')#0'#,)11$%H*+'1,0'1(&)=,+-,0%H5&%&'*'#, /?)'#&,H"+/&11&1@,/+%80'&,!9;.,0=&),C0(?,8)1&,"&7010+', )'=,1I1(&%,$10'#,)"#$%&'()*7&,1("$/($"&1,0',(?&,-+"%,+-, &OH5)')*+'1,-+","&7010+'1,+-,8&50&-,8)1&A,
  • 103. 4&50&-,6&7010+',0',!"#$%&'()*+', •  Q+C,8&50&-,"&7010+',(&/?'0G$&1,/)',8&,$1&=,0', )"#$%&'()*+',(?&+"IS, •  B+"<1,8I,s6+(1(&0't@,s;+#$055)'1<I,&(,)5At,)'=,s4+&55),&(,)5At, )%+'#,%+1(,/+%H"&?&'107&,)HH"+)/?&1,(+,)=="&11,), "&7010+',(?&+"I,-+",)"#$%&'(,1I1(&%1A, •  .&7&")5,C)I1,+-,)HH5I0'#,8&50&-,"&7010+',0',)"#$%&'()*+'>, –  F?)'#0'#,8I,)==0'#,+",=&5&*'#,)',)"#$%&'(A, –  F?)'#0'#,8I,)==0'#,+",=&5&*'#,),1&(,+-,)"#$%&'(1A, –  F?)'#0'#,(?&,)R)/<,D)'=a+",=&-&)(E,"&5)*+',)%+'#,)"#$%&'(1A, –  F?)'#0'#,(?&,1()($1,+-,8&50&-1,D)1,/+'/5$10+'1,+-,)"#$%&'(1EA, –  F?)'#0'#,(?&,(IH&,+-,)',)"#$%&'(,D-"+%,1("0/(,(+,=&-&)1085&@,+", 70/&,7&"1)EA,
  • 104. 9?"&&,)5(&"')*7&,)HH"+)/?&1, •  K)5)HH)@,M&"'U318&"'&"@,)'=,.0%)"0A,<")("A& -"9(3(/$I&JF0)%$%2/$3&%$;&K"A"%3(H)"& -"%3/$($EG,!"*`/0)5,3'(&550#&'/&,r+$"')5@, VYV>Vu[^@,[[A, •  F)I"+5@,=&,.)0'(,FI"@,)'=,p)#)1G$0&,./?0&OA, -"9(3(/$&/A&%$&.#E?5"$*%2/$&!L3*"5A,3',2"+/1, +-,M6,[^@,H)#&1,V[YuVXY@,[^A, •  4&'-&"?)(@,:$8+01@,)'=,2")=&A,=/>&*/&($A"#& A#/5&($7/$3(3*"$*&H")("A3&>(*1/?*&#"9(3($EA,3', 2"+/1,+-,3rF!3,V]]v@,H)#&1,VYY]uVYvv@,V]]vA,
  • 105. K)5)HH),&(,)5A,)HH"+)/?, •  !,<0'=,+-,'+'UH"0+"0*P&=,"&7010+',+H&")(+", 8)1&=,+',(?&,$1&,+-,&OH5)')*+'1A,, •  9?&,0=&),01,(?)(,)',)#&'(@,8&-+"&,0'/+"H+")*'#, 0'-+"%)*+',0'/+'101(&'(,C0(?,0(1,<'+C5&=#&@, "&G$&1(1,)',&OH5)')*+',1$HH+"*'#,0(A, •  F5)110/)5,=01*'/*+',8&(C&&'>, •  TOH5)')'=$%@,(?&,`')5,/+'/5$10+'@, •  TOH5)')'1@,1&(,+-,1&'(&'/&1,1$HH+"*'#,(?&, /+'/5$10+'A,
  • 106. 9?&,%+=&5, •  .("$/($"&,+-,&OH5)')*+',10%05)",(+,1("$/($"&,+-, =&=$/*7&,)"#$%&'(A, •  ,;)0',=0J&"&'/&>,, –  &7&"I,8&50&-,+-,)',&OH5)')*+',01,$'=&-&)1085&,D0',),%+%&'(, +-,*%&E@, –  1+%&,8&50&-1,+-,)',)"#$%&'(,%)I,8&,=&-&)1085&,+", (&'()*7&5I@,, –  &7&"I,&OH5)')*+',/+'()0'1,"$5&1,)'=,-)/($)5,<'+C5&=#&A, 3-,(?&,1&'(&'/&1,0',(?&,&OH5)')'1,)"&,8&R&",+",%+"&, H5)$1085&,(?)',(?&,1&'(&'/&1,0',(?&,+"0#0')5,8&50&-,8)1&@, (?&',(?&,&OH5)')*+',01,0'/+"H+")(&=A,
  • 107. 9?&,%+=&5, •  h+(,8&50&-1@,8$(,&OH5)')*+'1,D)'=,?&'/&, )"#$%&'(1E,1$HH+"*'#,),8&50&-,)"&,$1&=,-+", (?&,/?)'#&,H"+/&11A,, •  F+'10=&",8+(?,<&"'&5,)'=,H)"*)5,%&&(,"&7010+', 8I,),1&(,+-,1&'(&'/&1,)'=,#)7&,"&H"&1&'()*+', (?&+"&%1,-+",(?&%A,, •  9?&1&,+H&")(+"1,%)I,H)"*)55I,)//&H(,(?&,'&C, 0'-+"%)*+'@,1+,(?&I,)"&,'+'UH"0+"0*P&=A,
  • 108. 9?"&&,)5(&"')*7&,)HH"+)/?&1, •  K)5)HH)@,M&"'U318&"'&"@,)'=,.0%)"0A,<")("A& -"9(3(/$I&JF0)%$%2/$3&%$;&K"A"%3(H)"& -"%3/$($EA,!"*`/0)5,3'(&550#&'/&,r+$"')5@, VYV>Vu[^@,[[A, •  F)I"+5@,=&,.)0'(,FI"@,)'=,p)#)1G$0&,./?0&OA, -"9(3(/$&/A&%$&.#E?5"$*%2/$&!L3*"5A,3',2"+/1, +-,M6,[^@,H)#&1,V[YuVXY@,[^A, •  4&'-&"?)(@,:$8+01@,)'=,2")=&A,=/>&*/&($A"#& A#/5&($7/$3(3*"$*&H")("A3&>(*1/?*&#"9(3($EA,3', 2"+/1,+-,3rF!3,V]]v@,H)#&1,VYY]uVYvv@,V]]vA,
  • 109. F)I"+5,&(,)5A,)HH"+)/?, •  2"+H+1&,),:$'#U1(I5&,)81(")/(,)"#$%&'()*+', 1I1(&%,)55+C0'#,(?&,)==0*+',+-,),'&C, )"#$%&'(,C?0/?,%)I,0'(&")/(,C0(?,H"&70+$1, )"#$%&'(1A,, •  !"#$%&'()*+',-")%&C+"<,⟨!@6⟩,0=&'*`&=, C0(?,)',)11+/0)(&=,)R)/<,#")H?,gA, •  6&7010+',H"+/&11,H"+=$/&1,),'&C,-")%&C+"<, "&H"&1&'(&=,8I,),#")H?,gj,)'=,),'&C,1&(,+-, &O(&'10+'1A,
  • 110. 9?&,%+=&5, •  Q+C,(?&,1&(,+-,&O(&'10+'1,01,%+=0`&=,$'=&",(?&,"&7010+', H"+/&11S, •  9IH+5+#I,+-,=0J&"&'(,"&7010+'1>, –  4*-./.;*+B*;./.01>,+'5I,+'&,)//&H()85&,1&(,+-,)"#$%&'(1,0',(?&, "&701&=,-")%&C+"<@, –  *%(&1/.;*+B*;./.01>,)==1,(?&,'&C,)"#$%&'(,(+,(?&,&O01*'#, &O(&'10+'1@, –  /*)*-D;*+B*;./.01>,=&/"&)1&,+-,(?&,'$%8&",+-,/?+0/&1@, –  L=*/D01.15+B*;./.01>,")01&,)%80#$0(I@,8I,0'/"&)10'#,(?&,'$%8&", +-,&O(&'10+'1@, –  4*/9B=-D;*+B*;./.01>,"&%+70'#,&7&"I,&O(&'10+'@, –  &)9*B.15+B*;./.01>,1+%&,&O(&'10+'1,D)55,+-,(?&%E,)"&,)5(&"&=A,
  • 111. 9?"&&,)5(&"')*7&,)HH"+)/?&1, •  K)5)HH)@,M&"'U318&"'&"@,)'=,.0%)"0A,<")("A& -"9(3(/$I&JF0)%$%2/$3&%$;&K"A"%3(H)"& -"%3/$($EG,!"*`/0)5,3'(&550#&'/&,r+$"')5@, VYV>Vu[^@,[[A, •  F)I"+5@,=&,.)0'(,FI"@,)'=,p)#)1G$0&,./?0&OA, -"9(3(/$&/A&%$&.#E?5"$*%2/$&!L3*"5G,3',2"+/1, +-,M6,[^@,H)#&1,V[YuVXY@,[^A, •  4&'-&"?)(@,:$8+01@,)'=,2")=&A,=/>&*/&($A"#& A#/5&($7/$3(3*"$*&H")("A3&>(*1/?*&#"9(3($EG&3', 2"+/1,+-,3rF!3,V]]v@,H)#&1,VYY]uVYvv@,V]]vA,
  • 112. 4&'-&"?)(,&(,)5A,)HH"+)/?, •  ;+=&5,+"0&'(&=,(+C)"=1,(?&,("&)(%&'(,+-, 0'/+'101(&'/I,/)$1&=,8I,(?&,$1&,+-,%$5*H5&, 1+$"/&1,+-,0'-+"%)*+'A,, •  M'+C5&=#&,8)1&1,)"&,1(")*`&=>, –  &)/?,-+"%$5),0',(?&,M4,01,)11+/0)(&=,C0(?,0(1,5&7&5, +-,/&"()0'(I,/+""&1H+'=0'#,(+,(?&,5)I&",(+,C?0/?,0(, 8&5+'#1A,
  • 113. 9?&,%+=&5, •  9C+,/5)11&1,+-,)HH"+)/?&1,(+,=&)5,C0(?,0'/+'101(&'/I,0', M4>,/+?&"&'/&,(?&+"0&1,)'=,-+$'=)*+',(?&+"0&1A,, –  K0"1(,0'101(1,+',"&7010'#,(?&,M4,)'=,"&1(+"0'#,/+'101(&'/I@, –  p)R&",)//&H(1,0'/+'101(&'/I,)'=,/+H&1,C0(?,0(A, •  F+?&"&'/&,(?&+"0&1,H"+H+1&,, –  (+,#07&,$H,1+%&,-+"%$5)1,+-,(?&,M4,0',+"=&",(+,#&(,+'&,+", 1&7&")5,/+'101(&'(,1$8U8)1&1@, –  (+,)HH5I,/5)110/)5,&'()05%&'(,+',(?&1&,/+'101(&'(,1$8U8)1&1, (+,=&=$/&,H5)$1085&,/+'/5$10+'1,+-,(?&,M4A,, •  K+$'=)*+',(?&+"0&1,"&()0',, –  )55,)7)05)85&,0'-+"%)*+',, –  &)/?,H5)$1085&,/+'/5$10+',0'-&""&=,-"+%,(?&,M4,01,i$1*`&=, 8I,1("+'#,)"#$%&'()*7&,"&)1+'1,-+",8&50&70'#,0',0(A,
  • 114. 9?&,%+=&5, •  F5)0%>,, –  0(,=+&1,'+(,)5C)I1,%)<&,1&'1&,(+,"&701&,)', 0'/+'101(&'(,M4@,0',H)"*/$5)"@,0-,0'-+"%)*+',/+%&1, -"+%,%$5*H5&,1+$"/&1f, •  '+(,&7&','&/&11)"I,(+,"&1(+"&,/+'101(&'/I,0',+"=&", (+,%)<&,1&'1085&,0'-&"&'/&1,-"+%,)',0'/+'101(&'(, M4@,10'/&,0'-&"&'/&,8)1&=,+',)"#$%&'()*+',/)', =&"07&,/+'/5$10+'1,)'=,"&)1+'1,(+,8&50&7&,(?&%@, 0'=&H&'=&'(5I,+-,/+'101(&'/I,+-,(?&,M4f,
  • 116. 9"$1(,0',;!., •  ;&/?)'01%,-+",%)')#0'#,$'/&"()0'(I,)8+$(, )$(+'+%+$1,&'**&1,)'=,(?&,0'-+"%)*+',(?&I, 1(+"&A, •  !#&'(1,?)7&,(+,"&)1+',)8+$(,, •  )%+$'(,(?)(,(?&I,("$1(,(?+1&,+(?&",&'**&1@, •  C?&(?&",(?&I,)"&,("$1*'#,(?+1&,&'**&1,(+,/)""I, +$(,1+%&,()1<@,+",, •  C?&(?&",(?&I,)"&,("$1*'#,(?+1&,&'**&1,(+,'+(, %01$1&,/"$/0)5,0'-+"%)*+'A,
  • 117. 9"$1(, •  F)1(&5-")'/?0,)'=,K)5/+'&>,c%&5"$*%)&3*%*"I&%& 7/50)"F&%M*?;"&/A&%$&%E"$*&F&*/>%#;3& %$/*1"#&%E"$*&L&%H/?*&*1"&H"1%9(/?#N%72/$&%& #")"9%$*&A/#&*1"&E/%)&EdA, •  g)%8&R)>,c*#?3*&(3&*1"&3?HO"729"&0#/H%H()(*L& HL&>1(71&%$&($;(9(;?%)&.&"F0"7*3&*1%*&%$/*1"#& ($;(9(;?%)&<&0"#A/#53&%&E(9"$&%72/$&/$&>1(71& (*3&>")A%#"&;"0"$;3dA,,
  • 118. 9"$1(, •  p0)$>,c(A&%E"$*&(&H")("9"3&*1%*&%E"$*&O&1%3&*/);& 1(5&*1"&*#?*1&/$&0I&%$;&1"&*#?3*3&*1"& O?;E"5"$*&/A&O&/$&0I&*1"$&1"&>())&%)3/&H")("9"& 0dA,, •  F+%%+',&5&%&'(1,)"&, –  /+'101(&'(,=&#"&&,+-,$'/&"()0'(I@,)'=,, –  /+'b0/*'#,0'-+"%)*+',)11+/0)(&=,C0(?,("$1(A,,
  • 119. !"#$%&'()*+',)'=,9"$1(, •  M&I,)1H&/(,+-,("$1(>,0(,1(&%1,-"+%,(?&, "&5)*+'1?0H,8&(C&&',0'=070=$)51@,+",#"+$H1,+-, 0'=070=$)51A, •  9C+,%)i+",)1H&/(1,'&&=,(+,8&,?)'=5&=,8I,)'I, "&H"&1&'()*+',+-,("$1(>, –  %&)1$"&1,+-,("$1(@,)'=, –  H"+7&')'/&,+-,("$1(,0'-+"%)*+'A,
  • 120. !"#$%&'()*+',)'=,9"$1(, •  9"$1(,`&5=,!,%&(?+=+5+#0&1,-+",/+%H$*'#, ("$1(A, •  !"#$%&'()*+',`&5=,!,%&(?+=+5+#0&1,-+", "&)1+'0'#,)8+$(,("$1(A, !"#$%&'()*+',/)',H"+70=&,),%&/?)'01%,-+", ?)'=50'#,%)'I,+-,(?&,)1H&/(1,(?)(,C&,'&&=,(+, /)H($"&,)8+$(,("$1(A,
  • 121. !"#$%&'()*+',)'=,9"$1(, •  !81(")/(,)HH"+)/?&1,1$/?,)1,:$'#,)'=,0(1, =&"07)*7&1@,("&)(,)"#$%&'(1,)1,)(+%0/,+8i&/(1A, •  p0R5&,+",'+(?0'#,)8+$(,0'(&"')5,1("$/($"&,+-, )"#$%&'(A, •  h+,%&/?)'01%,(+,"&H"&1&'(,(?&,1+$"/&,+-,(?&, 0'-+"%)*+',-"+%,C?0/?,(?&,)"#$%&'(,01, /+'1("$/(&=A, •  F)''+(,/)H($"&>, •  ),)R)/<1,8,8&/)$1&,8,01,8)1&=,+',0'-+"%)*+',-"+%, 1+$"/&,1@,)'=,(?&"&,01,&70=&'/&,(?)(,1+$"/&,1,01,'+(, ("$1(C+"(?IA,
  • 122. !"#$%&'()*+',)'=,9"$1(, •  !'+(?&",H"+85&%,C0(?,:$'#j1,)"#$%&'()*+', 1I1(&%,-"+%,(?&,H&"1H&/*7&,+-,"&)1+'0'#, )8+$(,("$1(,01, –  0(,?)1,'+,&OH50/0(,%&)'1,(+,"&H"&1&'(,=&#"&&1,+-, ("$1(A,, •  F"0*/)5,H+0'(>,C?&(?&",1+%&+'&j1,)"#$%&'(,01, ("$1(C+"(?I,+",'+(@,(?&,H"&7)5&'/&,+-, '$%&"0/)5,%&)1$"&1,+-,("$1(,0',(?&,50(&")($"&, 5&)=1,$1,(+,C)'(,(+,"&H"&1&'(,(?&1&A,
  • 123. !"#$%&'()*+',)'=,9"$1(, •  L(?&",0%H+"()'(,011$&1>, –  K&&=8)/<,-"+%,)"#$%&'(1,(+,1+$"/&1@,)'=,8)/<A, –  :0J&"&'(,H&"1H&/*7&1,+',("$1(>,, •  10'/&"0(I,k.,/+%H&(&'/&, –  :I')%0/1,+-,("$1(,)51+,=$&,(+,(?&,)R)/<1,H"&1&'(, 0',(?&,)"#$%&'()*+',-")%&C+"<1A,
  • 124. !"#$%&'()*+',)'=,9"$1(, •  3%H+"()'(,)1H&/(,+-,"&)1+'0'#,)8+$(,("$1(>, '&&=,-+",)',0'=070=$)5,(+,8&,)85&,(+,"&701&,(?&, ("$1(,?&,?)1,0',)'+(?&",8)1&=,+',&OH&"0&'/&A,, •  9?01,)55+C1,(+,"&H"&1&'(,(?&,/)1&,0',C?0/?,+'&, 0'=070=$)5,"&701&1,0(1,70&C,+-,),1+$"/&,)1,), "&1$5(,+-,/+'10=&"0'#,0'-+"%)*+',H"+70=&=,8I, )'+(?&",0'=070=$)5A,
  • 125. 9?"&&,)5(&"')*7&,)HH"+)/?&1, •  2)"1+'1@,9)'#@,.<5)"@,F)0@,;/4$"'&IA, .#E?5"$*%2/$PH%3";&#"%3/$($E&($&%E"$*3&>(*1& 9%#L($E&;"E#""3&/A&*#?3*G&3',2"+/1,+-,!!;!.,[VVA, •  =),F+1(),2&"&0")@,9&R)%)'P0@,k055)()A,F?)'#0'#, +'&j1,;0'=>,T")1&,+",6&C0'=S,2+11080501*/,4&50&-, 6&7010+',C0(?,K$PPI,!"#$%&'()*+',4)1&=,+', 9"$1(A,3',2"+/1,+-,3rF!3,[VV@,H)#&1,VZYUVWV@, [VVA, •  ;)R@,;+"#&@,9+'0A,Q/5H($($E&3*%23273&%$;& %#E?5"$*3&*/&7/50?*"&*#?3*G,3',2"+/1,+-,!!;!., [V@,H)#&1,[]U[VZ@,[VA,
  • 126. 2)"1+'1,&(,)5A,)HH"+)/?, •  !11$%&,(?)(,)#&'(1,%)0'()0',),("$1(,'&(C+"<,+-, (?&0",)/G$)0'()'/&1@,C?0/?,0'/5$=&1,, –  ")*'#1,+-,?+C,%$/?,(?+1&,)/G$)0'()'/&1,)"&,("$1(&=@, –  ?+C,%$/?,(?+1&,)/G$)0'()'/&1,("$1(,(?&0", )/G$)0'()'/&1@,)'=,1+,+'A, •  B?)(,0'-&"&'/&,01,"&)1+')85&,0',1$/?,'&(C+"<1S, •  2"+H)#)*+',+-,("$1(,)'=,H"+7&')'/&,w,8+(?,(?&, (")'10*70(I,+-,("$1(,"&5)*+'1,)'=,%+"&,/+%H5&O, "&5)*+'1?0H1,50<&,c/+U/0()*+'jjA,
  • 127. 9?&,-")%&C+"<, •  K")%&C+"<,(+,0'("+=$/&,(?&,1+$"/&1,0',)"#$%&'()*+'@, )'=,(+,&OH50/0(5I,&OH"&11,=&#"&&1,+-,("$1(A,, •  !55,)#&'(1,(?)(,?)7&,/+%%$'0/)(&=,0'-+"%)*+',(+,!#, )"&,%&%8&"1,+-,!#j1,1+/0)5,'&(C+"<, •  3(,01,H+11085&,, –  (+,/+'1("$/(,),#")H?,C?0/?,"&5)(&1,!#,(+,)55,(?&1&,)#&'(1f,, –  (+,)R)/?,),'$%&"0/)5,%&)1$"&,(+,&)/?,50'<,0',(?01,1+/0)5, '&(C+"<,(+,G$)'*-I,(?&,&O(&'(,(+,C?0/?,)',)#&'(,("$1(1, (?+1&,(+,C?0/?,0(,01,50'<&=,0',(?&,1+/0)5,'&(C+"<f, •  .("$/($"&,/)55&=,("$1(,'&(C+"<A,
  • 128. 9?&,-")%&C+"<, •  2"++-,("&&1,C?&"&,/+'/5$10+'1,+-,(?&,("&&1,)"&, -+"%$5)&,+-,0'(&"&1(A, •  T)/?,-+"%$5),(?)(,01,'+(,)',0'-&"&'/&,01,50'<&=,(+,(?&, )#&'(,1$HH5I0'#,(?&,0'-+"%)*+'A,, •  TA#A@,)"#$%&'(1,/+'/&"',C?&(?&",+",'+(,(+,C)(/?, !5%+=+7)",`5%,cQ)85&,/+',&55)dA, •  4&50&-,7)5$&1,-+",0'-+"%)*+',(?)(,/+%&1,-"+%,=0J&"&'(, )#&'(1,1A(A,=&H&'=1,+',("$1(,0',(?&,1+$"/&,)#&'(D1EA,, •  !"/1,0=&'*-I,/+'b0/(1,8&(C&&',)"#$%&'(1A,, •  9?&1&,/+%H+'&'(1,%)<&,$H,),("$1(U&O(&'=&=, )"#$%&'()*+',#")H?A,
  • 131. 9?"&&,)5(&"')*7&,)HH"+)/?&1, •  2)"1+'1@,9)'#@,.<5)"@,F)0@,;/4$"'&IA, .#E?5"$*%2/$PH%3";&#"%3/$($E&($&%E"$*3&>(*1& 9%#L($E&;"E#""3&/A&*#?3*G&3',2"+/1,+-,!!;!.,[VVA, •  =),F+1(),2&"&0")@,9&R)%)'P0@,k055)()A,F?)'#0'#, +'&j1,;0'=>,T")1&,+",6&C0'=S,2+11080501*/,4&50&-, 6&7010+',C0(?,K$PPI,!"#$%&'()*+',4)1&=,+', 9"$1(A,3',2"+/1,+-,3rF!3,[VV@,H)#&1,VZYUVWV@, [VVA, •  ;)R@,;+"#&@,9+'0A,Q/5H($($E&3*%23273&%$;& %#E?5"$*3&*/&7/50?*"&*#?3*G,3',2"+/1,+-,!!;!., [V@,H)#&1,[]U[VZ@,[VA,
  • 132. =),F+1(),2&"&0"),&(,)5A,)HH"+)/?, , •  p+11,+-,0'-+"%)*+',0',/)1&,+-,"&0'1()(&%&'(,+-, H"&70+$1,0'-+"%)*+'A, •  2"0'/0H5&,+-,cH"0+"0(I,(+,0'/+%0'#,0'-+"%)*+'d, •  :")C8)/<1,0',%$5*)#&'(,1I1(&%1>, –  /?"+'+5+#0/)5,1&G$&'/&,+-,)""07)5,+-,0'-+"%)*+'@, –  '+(?0'#,(+,=+,C0(?,(?&0",("$1()8050(IA,
  • 133. 9?&,%+=&5 , •  !"#$%&'(1,⟨x@,y⟩,1$HH+"(,)#&'(1j,8&50&-1@, •  4&50&-1,)"&,/+'/5$10+'1,+-,)"#$%&'(1@, •  9"$1(C+"(?0'&11,%&)1$"&=,8I,$10'#, H"+8)8050*&1@4N9,+'5I,0-,=)(),)7)05)85&@, •  2+11080501*/,5+#0/,C&55,1$0(&=,(+,=&)5,C0(?, 0'/+%H5&(&,0'-+"%)*+'A,
  • 134. 9?&,%+=&5 , •  K$PPI,&7)5$)*+',+-,(?&,)"#$%&'(1@,, •  !"#$%&'(1,)11+/0)(&=,C0(?,=&#"&&,+-, H5)$108050(I@,, •  9"$1(C+"(?0'&11,+-,(?&,1+$"/&,+-,0'-+"%)*+', •  !"#$%&'(1,&7)5$)(&=,0',#")=$)5,C)I, =&H&'=0'#,+',=&#"&&,+-,("$1(A,
  • 135. 9?&,%+=&5 , '(#")*$ +,-.,/' ! ! " !"#$% & " 0 !
  • 136. TO)%H5&, B I saw John killing Mary, thus John killed Mary. A Wit1 If John did not kill Mary, then John is innocent. Judge D Mary was killed before 6 p.m., thus when Mary was killed the show was still to begin C John was at the theater with Corner me when Mary was killed, thus John did not kill Mary. Wit2
  • 137. TO)%H5&, B A(B) = 0.2 I saw John killing Mary, A(A) = 1.0 thus John killed Mary. A Wit1 If John did not kill Mary, then John is innocent. Judge A(D) = 0.3 D Mary was killed before 6 p.m., thus when Mary was killed the show was still to begin C John was at the theater with Corner me when Mary was killed, thus John did not kill Mary. Wit2 A(C) = 1.0
  • 138. TO)%H5&, B α(B) = 0.2 I saw John killing Mary, α(A) = 0.8 thus John killed Mary. A Wit1 If John did not kill Mary, then John is innocent. Judge α(D) = 0.3 D Mary was killed before 6 p.m., thus when Mary was killed the show was still to begin C John was at the theater with Corner me when Mary was killed, thus John did not kill Mary. Wit2 α(C) = 0.7
  • 139. 9?"&&,)5(&"')*7&,)HH"+)/?&1, •  2)"1+'1@,9)'#@,.<5)"@,F)0@,;/4$"'&IA, .#E?5"$*%2/$PH%3";&#"%3/$($E&($&%E"$*3&>(*1& 9%#L($E&;"E#""3&/A&*#?3*G&3',2"+/1,+-,!!;!.,[VVA, •  =),F+1(),2&"&0")@,9&R)%)'P0@,k055)()A,F?)'#0'#, +'&j1,;0'=>,T")1&,+",6&C0'=S,2+11080501*/,4&50&-, 6&7010+',C0(?,K$PPI,!"#$%&'()*+',4)1&=,+', 9"$1(A,3',2"+/1,+-,3rF!3,[VV@,H)#&1,VZYUVWV@, [VVA, •  ;)R@,;+"#&@,9+'0A,Q/5H($($E&3*%23273&%$;& %#E?5"$*3&*/&7/50?*"&*#?3*G,3',2"+/1,+-,!!;!., [V@,H)#&1,[]U[VZ@,[VA,
  • 140. ;)R,&(,)5A,)HH"+)/?, •  .()"*'#,-"+%,z$,)'=,.0'#?>,)HH"+)/?,(+,("$1(, $10'#,:&%H1(&"U.?)-&",8&50&-,-$'/*+',=&"07&=, -"+%,1()*1*/)5,=)(),/+'/&"'0'#,()"#&(j1, 8&?)70+$"A, •  TO(&'10+',+-,z$,)'=,.0'#?j1,)HH"+)/?,8I, )55+C0'#,&7)5$)(+",(+,()<&,0'(+,)//+$'(@,0', )==0*+',(+,1()*1*/)5,=)()@,i$1*`&=,/5)0%1, /+'/&"'0'#,&OH&/(&=,8&?)70+$",+-,(?&,()"#&(A,
  • 141. 9?&,%+=&5, •  F5)0%1,-+"%,8)101,+-,&7)5$)(+"j1,+H0'0+'1A, •  K+"%)55I,"&H"&1&'(&=,8I,)"#$%&'(1,0',)81(")/(, )"#$%&'()*+'A, •  9C+,/5)11&1,+-,)"#$%&'(1>,, –  >0B*-&/9+&B5='*19/@,0',-)7+$",+",)#)0'1(,("$1*'#, (?&,()"#&(@,)'=,, –  '.D5&D01+&B5='*19/@,)R)/<0'#,-+"&/)1(, )"#$%&'(1,+",+(?&",%0*#)*+',)"#$%&'(1A,
  • 142. 9?&,%+=&5, •  ;&(?+=,-+",/+'1("$/*'#,:&%H1(&"U.?)-&", 8&50&-,-$'/*+'1,-"+%,1()*1*/)5,=)(),)'=,(?&1&, )"#$%&'(1A, •  h&C,)"#$%&'()*+'U8)1&=,8&50&-,-$'/*+',0', (&"%1,+-,'&C,)"#$%&'()*+'U8)1&=,&70=&'/&, %)11,-$'/*+',%l),/+%80'0'#,1()*1*/)5, &70=&'/&,)'=,)"#$%&'(1,)1,&70=&'/&A,
  • 143. 9?&,%+=&5, •  K+",&)/?,=0%&'10+',=@,/+'10=&",)',)81(")/(,!K,K=, /+'101*'#,+-,(?&,-+55+C0'#,)"#$%&'(1>, –  K+"&/)1(,)"#$%&'(,e(,1$HH+"*'#,e9,D0A&A,{)De(E,|,e9E, +',(?&,#"+$'=,(?)(,(?&"&,01,'+,#$)")'(&&,0',(?&,-+"%, +-,C"0R&',/+'(")/(,/5)$1&,/+'/&"'0'#,=f, –  K+"&/)1(,)"#$%&'(,(,1$HH+"*'#,9,D0A&A,{),D(E|9E,+',(?&, #"+$'=,(?)(,(?&"&,&O01(1,),#$)")'(&&,0',(?&,-+"%,+-,), /+'(")/(,/5)$1&,/+'/&"'0'#,=f, –  ;0*#)*+',)"#$%&'(,7,)R)/<0'#,(,+',(?&,#"+$'=,(?)(, (?&,()"#&(,?)1,0',(?&,H)1(,c%+1(,+}&'d,70+5)(&=, &O01*'#,/+'(")/(,/5)$1&1,/+'/&"'0'#,=A,
  • 144. 9?&,%+=&5, •  K+",&)/?,=0%&'10+',=@,, –  &0(?&",(?&"&,&O01(1,),/+'(")/(,/5)$1&,/+'/&"'0'#,=,)'=, (,8&5+'#1,(+,D1&(,+-,)"#$%&'(1,0'E,K=@,, –  +",(?&"&,=+&1,'+(,&O01(,1$/?,),/5)$1&@,)'=,e(,8&5+'#1, (+,K=A,, •  3-,(?&"&,01,),/+'(")/(,/5)$1&@,)'=,0(,?)1,8&&', +81&"7&=,(?)(@,0',(?&,H)1(@,C0(?,)',)')5+#+$1, /5)$1&,(?&,0'(&")/*+',C0(?,()"#&(,=0=,'+(,&O?080(, =,)(,)//&H()85&,5&7&5,0',),%)i+"0(I,+-,/)1&1>,, –  K=,/+'()0'1,7,)1,C&55,)1,)',)R)/<,-"+%,7,)#)0'1(,(A,