<s>
Affinity	B-Algorithm
analysis	I-Algorithm
falls	O
under	O
the	O
umbrella	O
term	O
of	O
data	B-Application
mining	I-Application
which	O
uncovers	O
meaningful	O
correlations	O
between	O
different	O
entities	O
according	O
to	O
their	O
co-occurrence	O
in	O
a	O
data	O
set	O
.	O
</s>
<s>
In	O
almost	O
all	O
systems	O
and	O
processes	O
,	O
the	O
application	O
of	O
affinity	B-Algorithm
analysis	I-Algorithm
can	O
extract	O
significant	O
knowledge	O
about	O
the	O
unexpected	O
trends	O
.	O
</s>
<s>
In	O
fact	O
,	O
affinity	B-Algorithm
analysis	I-Algorithm
takes	O
advantages	O
of	O
studying	O
attributes	O
that	O
go	O
together	O
which	O
helps	O
uncover	O
the	O
hidden	O
pattens	O
in	O
a	O
big	O
data	O
through	O
generating	O
association	B-Algorithm
rules	I-Algorithm
.	O
</s>
<s>
Association	B-Algorithm
rules	I-Algorithm
mining	O
procedure	O
is	O
two-fold	O
:	O
first	O
,	O
it	O
finds	O
all	O
frequent	O
attributes	O
in	O
a	O
data	O
set	O
and	O
,	O
then	O
generates	O
association	B-Algorithm
rules	I-Algorithm
satisfying	O
some	O
predefined	O
criteria	O
,	O
support	O
and	O
confidence	O
,	O
to	O
identify	O
the	O
most	O
important	O
relationships	O
in	O
the	O
frequent	O
itemset	O
.	O
</s>
<s>
The	O
association	B-Algorithm
rules	I-Algorithm
mining	O
takes	O
the	O
form	O
of	O
if	O
a	O
condition	O
or	O
feature	O
(	O
A	O
)	O
is	O
present	O
then	O
another	O
condition	O
or	O
feature	O
(	O
B	O
)	O
exists	O
.	O
</s>
<s>
There	O
are	O
two	O
important	O
metrics	O
for	O
performing	O
the	O
association	B-Algorithm
rules	I-Algorithm
mining	O
technique	O
:	O
support	O
and	O
confidence	O
.	O
</s>
<s>
Also	O
,	O
a	B-Algorithm
priori	I-Algorithm
algorithm	I-Algorithm
is	O
used	O
to	O
reduce	O
the	O
search	O
space	O
for	O
the	O
problem	O
.	O
</s>
<s>
The	O
support	O
metric	O
in	O
the	O
association	B-Algorithm
rule	I-Algorithm
learning	I-Algorithm
algorithm	O
is	O
defined	O
as	O
the	O
frequency	O
of	O
the	O
antecedent	O
or	O
consequent	O
appearing	O
together	O
in	O
a	O
data	O
set	O
.	O
</s>
<s>
Moreover	O
,	O
confidence	O
is	O
expressed	O
as	O
the	O
reliability	O
of	O
the	O
association	B-Algorithm
rules	I-Algorithm
determined	O
by	O
the	O
ratio	O
of	O
the	O
data	O
records	O
containing	O
both	O
A	O
and	O
B	O
.	O
</s>
<s>
Considering	O
all	O
the	O
above-mentioned	O
definitions	O
,	O
affinity	B-Algorithm
analysis	I-Algorithm
can	O
develop	O
rules	O
that	O
will	O
predict	O
the	O
occurrence	O
of	O
an	O
event	O
based	O
on	O
the	O
occurrence	O
of	O
other	O
events	O
.	O
</s>
<s>
This	O
data	B-Application
mining	I-Application
method	O
has	O
been	O
explored	O
in	O
different	O
fields	O
including	O
disease	O
diagnosis	O
,	O
market	B-Algorithm
basket	I-Algorithm
analysis	I-Algorithm
,	O
retail	O
industry	O
,	O
higher	O
education	O
,	O
and	O
financial	O
analysis	O
.	O
</s>
<s>
In	O
retail	O
,	O
affinity	B-Algorithm
analysis	I-Algorithm
is	O
used	O
to	O
perform	O
market	B-Algorithm
basket	I-Algorithm
analysis	I-Algorithm
,	O
in	O
which	O
retailers	O
seek	O
to	O
understand	O
the	O
purchase	O
behavior	O
of	O
customers	O
.	O
</s>
<s>
Market	B-Algorithm
basket	I-Algorithm
analysis	I-Algorithm
might	O
tell	O
a	O
retailer	O
that	O
customers	O
often	O
purchase	O
shampoo	O
and	O
conditioner	O
together	O
,	O
so	O
putting	O
both	O
items	O
on	O
promotion	O
at	O
the	O
same	O
time	O
would	O
not	O
create	O
a	O
significant	O
increase	O
in	O
revenue	O
,	O
while	O
a	O
promotion	O
involving	O
just	O
one	O
of	O
the	O
items	O
would	O
likely	O
drive	O
sales	O
of	O
the	O
other	O
.	O
</s>
<s>
Market	B-Algorithm
basket	I-Algorithm
analysis	I-Algorithm
may	O
provide	O
the	O
retailer	O
with	O
information	O
to	O
understand	O
the	O
purchase	O
behavior	O
of	O
a	O
buyer	O
.	O
</s>
<s>
This	O
kind	O
of	O
analysis	O
is	O
supposedly	O
an	O
example	O
of	O
the	O
use	O
of	O
data	B-Application
mining	I-Application
.	O
</s>
<s>
A	O
widely	O
used	O
example	O
of	O
cross	O
selling	O
on	O
the	O
web	O
with	O
market	B-Algorithm
basket	I-Algorithm
analysis	I-Algorithm
is	O
com	O
's	O
use	O
of	O
"	O
customers	O
who	O
bought	O
book	O
A	O
also	O
bought	O
book	O
B	O
"	O
,	O
e.g.	O
</s>
<s>
Market	B-Algorithm
basket	I-Algorithm
analysis	I-Algorithm
can	O
be	O
used	O
to	O
divide	B-Algorithm
customers	I-Algorithm
into	I-Algorithm
groups	I-Algorithm
.	O
</s>
<s>
Business	O
use	O
of	O
market	B-Algorithm
basket	I-Algorithm
analysis	I-Algorithm
has	O
significantly	O
increased	O
since	O
the	O
introduction	O
of	O
electronic	O
point	O
of	O
sale	O
.	O
</s>
<s>
Amazon	B-Application
uses	O
affinity	B-Algorithm
analysis	I-Algorithm
for	O
cross-selling	O
when	O
it	O
recommends	O
products	O
to	O
people	O
based	O
on	O
their	O
purchase	O
history	O
and	O
the	O
purchase	O
history	O
of	O
other	O
people	O
who	O
bought	O
the	O
same	O
item	O
.	O
</s>
<s>
Family	O
Dollar	O
plans	O
to	O
use	O
market	B-Algorithm
basket	I-Algorithm
analysis	I-Algorithm
to	O
help	O
maintain	O
sales	O
growth	O
while	O
moving	O
towards	O
stocking	O
more	O
low-margin	O
consumable	O
goods	O
.	O
</s>
<s>
An	O
important	O
clinical	O
application	O
of	O
affinity	B-Algorithm
analysis	I-Algorithm
is	O
that	O
it	O
can	O
be	O
performed	O
on	O
medical	O
patient	O
records	O
in	O
order	O
to	O
generate	O
association	B-Algorithm
rules	I-Algorithm
.	O
</s>
<s>
The	O
obtained	O
association	B-Algorithm
rules	I-Algorithm
can	O
be	O
further	O
assessed	O
to	O
find	O
different	O
conditions	O
and	O
features	O
that	O
coincide	O
on	O
a	O
large	O
block	O
of	O
information	O
.	O
</s>
<s>
In	O
addition	O
to	O
exploring	O
the	O
association	O
between	O
different	O
symptoms	O
in	O
a	O
patient	O
related	O
to	O
a	O
specific	O
disease	O
,	O
the	O
possible	O
correlations	O
between	O
various	O
diseases	O
contributing	O
to	O
another	O
condition	O
can	O
also	O
be	O
identified	O
using	O
affinity	B-Algorithm
analysis	I-Algorithm
.	O
</s>
