Scribd
Upload
46 views3 pages

CS 2742 (Logic in Computer Science) - Fall 2008: Antonina Kolokolova

1) The document discusses binary relations and their properties such as reflexive, symmetric, transitive, and equivalence relations. It provides examples like the parent relation and equality. 2) Transitive closure is introduced as a way to define relations like ancestry that are implied by other relations but not directly stated. 3) Modular arithmetic and congruence relations are discussed in the context of cryptography methods like Caesar ciphers. Wraparound is achieved using modulo operations.

Uploaded by

ZhichaoWang
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
46 views3 pages

CS 2742 (Logic in Computer Science) - Fall 2008: Antonina Kolokolova

1) The document discusses binary relations and their properties such as reflexive, symmetric, transitive, and equivalence relations. It provides examples like the parent relation and equality. 2) Transitive closure is introduced as a way to define relations like ancestry that are implied by other relations but not directly stated. 3) Modular arithmetic and congruence relations are discussed in the context of cryptography methods like Caesar ciphers. Wraparound is achieved using modulo operations.

Uploaded by

ZhichaoWang
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

CS 2742 (Logic in Computer Science) Fall 2008

Lecture 20
Antonina Kolokolova
November 10, 2008

7.1

Relations

Recall that a relation on n variables is just a subset of a Cartesian products of n sets which
are domains of the variables. In this respect, a binary relation on a set A is a just a subset
of A A.
For example, R(x, y) Z Z such that R(x, y) if and only if x y is a binary relation.
In this case, we often write directly x y, rather than writing R(x, y). The relation x = y
is also a binary relation, which can be defined for many different domains such as integers,
reals, strings and so on.
Another common binary relation is congruence mod n for a natural number n: in that
case, R(x, y), written as x y mod n, if z Z such that x y = zn. This is the same
as saying that x and y have the same remainder from division by n. For every n there is a
different mod n relation.
Yet another binary relation is P arent(x, y), which contains all pairs x, y such that x is a
parent of y. At this point you may ask what is the difference between a predicate P arent(x, y)
and a binary relation P arent(x, y): the predicate is true on the pairs (x, y) that belong to
the set which is the binary relation.
There are several major types of binary relations. In this case, it is often more convenient
to view a binary relation R(x, y) as taking a x into y. Binary relations can be:
Reflexive: x A R(x, x).
For example, x = y, x y, x y mod n are reflexive, but P arent9(x, y) and x < y
are not.
Symmetric: x, y A R(x, y) R(y, x).
57

Antisymmetric: x, y A R(x, y) R(y, x) x = y.


For example, x = y, x y mod n are symmetric, x y, x < y and P arent(x, y)
are antisymmetric, and the relation Likes(x, y) defined on the domain of people is
neither symmetric nor antisymmetric. Note that symmetric and antisymmetric are
not complements: an easy way to see that is by noticing that both have a universal
quantifier in the definition, whereas a complement of a universal quantifier is existential
quantifier. Another way of checking it is to find a relation such as Likes(x, y) which
is neither symmetric nor antisymmetric.
Transitive: x, y, z A R(x, y) R(y, z) R(x, z)
For example, relation x = y is transitive, because if x = y and y = z then x = z as
well. Same can be said about x y and x < y: if x is smaller than y, and y is smaller
than z, then x is definitely smaller than z.
Note that P arent(x, y) relation is not transitive: if x is a parent of y, and y is a parent
of z, then x is not a parent of z, it is a grandparent. But often we do want to express
the fact that one person is related to another by a chain of Parent(x, y) relationships:
that is, it is a grandparent, or great-grandparent, or great-great-grandparent and so
on. For that, we can define a relation Ancestor(x, y), which would contain all pairs
(x, y) related by a chain of P arent() relations. Such a relation is called a transitive
closure.
Definition 1. For a relation R(x, y), its transitive closure contains all pairs x, y such
that there is a sequence z1 , . . . , zn with R(x, z1 ), R(zn , y) and 1 i < n R(zi , zi+1 ).
For example, if John is a grandparent of Jill via Mary who is the daughter of John and
the mother of Jill, then x=John, y=Jill and n = 1, so there is just one z1 =Mary such
that Parent(John,Mary) and Parent(Mary, Jill). Thus, Ancestor(John, Mary) holds.
A way to describe the Ancestor relation is by the following recursive definition (of the
Ancestor predicate): Ancestor(x, y) = P arent(x, y) zP arent(x, z) Ancestor(z, y).
However, this definition is not a first-order formula it mentions the relation Ancestor(x, y)
which is being defined. Moreover, as we have seen with the Flight example in the predicate logic, transitive closure cannot be defined by a first-order formula.
A relation that is reflexive, symmetric and transitive is called an equivalence relation. Equality and congruence mod 2 are equivalences.So is equivalence of digital circuits computing the
same function. Any equivalence relation breaks up the set of all objects on which it is defined
into equivalence classes: in such a class all objects are equivalent to each other, but not to
any object outside of the class. For example, x = y mod 5 breaks all natural numbers into
5 equivalence classes: 1) numbers that are divisible by 5, 2) numbers that have a remainder
1 from division by 5, 3) ones with remainder 2, 4) with remainder 3, 4) with remainder 4. In
the first class, there are numbers 0,5,10,15, 20.., in the third 2,7,12,17,22,... It is possible
to extend this definition to all integers: -1 will be in the same class as 4, -2 in the same class
58

as 3, -5 as 5 and so on. For the names of the equivalence classes we can pick any element
of them, but it is convenient to choose 0, 1, 2, 3, 4, and in general, we represent equivalence
classes mod n as 0, 1, . . . , n 1.
There is a special term to denote relations which are reflexive, transitive and antisymmetric:
they are called partial order relations (e.g., subset relation). A total order is a subclass of
partial orders with an additional property that any two elements are related: that is, for any
x, y either R(x,y) or R(y,x). E.g.: on numbers.

7.2

Application to cryptography

Lets look more at the congruence modulo n relation.


Enumerate all letters: A=1, B=2, C=3, D=4,E=5,F=6, . . . , Z=26. In Caesars cipher,
a letter gets encoded by a letter which is 3 places further in the alphabet. For example,
the letter B gets encoded as E, A as D and so on. So the formula seems to be Code(i) =
Letter(i + 3). But this formula does not tell us how to encode, say, letter Z. Intuitively, we
want to wrap around and encode Z as C. How do we represent it by a formula? Here is
where mod n relation is useful. In order to represent wrap around, it is enough to write
Code(i) = Letter(i + 3) mod 26, where 26 is the number of letters. Now, to decode we just
take Letter(i) = Code(i 3) mod 26 (remember that our definition of mod can handle
negative numbers).
The problem with Caesars cipher is that seeing enough messages one can figure out the
code, and knowing how to encode can easily decode. In the next lecture we will see some
more secure cryptosystems.

59

You might also like