<s>
Data	B-Application
mining	I-Application
is	O
the	O
process	O
of	O
extracting	O
and	O
discovering	O
patterns	O
in	O
large	O
data	B-General_Concept
sets	I-General_Concept
involving	O
methods	O
at	O
the	O
intersection	O
of	O
machine	O
learning	O
,	O
statistics	O
,	O
and	O
database	B-Application
systems	I-Application
.	O
</s>
<s>
Data	B-Application
mining	I-Application
is	O
an	O
interdisciplinary	O
subfield	O
of	O
computer	B-General_Concept
science	I-General_Concept
and	O
statistics	O
with	O
an	O
overall	O
goal	O
of	O
extracting	O
information	O
(	O
with	O
intelligent	O
methods	O
)	O
from	O
a	O
data	B-General_Concept
set	I-General_Concept
and	O
transforming	O
the	O
information	O
into	O
a	O
comprehensible	O
structure	O
for	O
further	O
use	O
.	O
</s>
<s>
Data	B-Application
mining	I-Application
is	O
the	O
analysis	O
step	O
of	O
the	O
"	O
knowledge	B-Application
discovery	I-Application
in	I-Application
databases	I-Application
"	O
process	O
,	O
or	O
KDD	O
.	O
</s>
<s>
Aside	O
from	O
the	O
raw	O
analysis	O
step	O
,	O
it	O
also	O
involves	O
database	O
and	O
data	B-General_Concept
management	I-General_Concept
aspects	O
,	O
data	B-General_Concept
pre-processing	I-General_Concept
,	O
model	O
and	O
inference	O
considerations	O
,	O
interestingness	O
metrics	O
,	O
complexity	O
considerations	O
,	O
post-processing	O
of	O
discovered	O
structures	O
,	O
visualization	B-Application
,	O
and	O
online	B-Algorithm
updating	I-Algorithm
.	O
</s>
<s>
The	O
term	O
"	O
data	B-Application
mining	I-Application
"	O
is	O
a	O
misnomer	O
because	O
the	O
goal	O
is	O
the	O
extraction	B-General_Concept
of	O
patterns	O
and	O
knowledge	O
from	O
large	O
amounts	O
of	O
data	O
,	O
not	O
the	O
extraction	B-General_Concept
(	O
mining	O
)	O
of	O
data	O
itself	O
.	O
</s>
<s>
It	O
also	O
is	O
a	O
buzzword	O
and	O
is	O
frequently	O
applied	O
to	O
any	O
form	O
of	O
large-scale	O
data	O
or	O
information	O
processing	O
(	O
collection	O
,	O
extraction	B-General_Concept
,	O
warehousing	O
,	O
analysis	O
,	O
and	O
statistics	O
)	O
as	O
well	O
as	O
any	O
application	O
of	O
computer	B-Application
decision	I-Application
support	I-Application
system	I-Application
,	O
including	O
artificial	B-Application
intelligence	I-Application
(	O
e.g.	O
,	O
machine	O
learning	O
)	O
and	O
business	B-General_Concept
intelligence	I-General_Concept
.	O
</s>
<s>
The	O
book	O
Data	B-Application
mining	I-Application
:	O
Practical	O
machine	O
learning	O
tools	O
and	O
techniques	O
with	O
Java	B-Language
(	O
which	O
covers	O
mostly	O
machine	O
learning	O
material	O
)	O
was	O
originally	O
to	O
be	O
named	O
Practical	O
machine	O
learning	O
,	O
and	O
the	O
term	O
data	B-Application
mining	I-Application
was	O
only	O
added	O
for	O
marketing	O
reasons	O
.	O
</s>
<s>
Often	O
the	O
more	O
general	O
terms	O
(	O
large	O
scale	O
)	O
data	B-General_Concept
analysis	I-General_Concept
and	O
analytics	O
—	O
or	O
,	O
when	O
referring	O
to	O
actual	O
methods	O
,	O
artificial	B-Application
intelligence	I-Application
and	O
machine	O
learning	O
—	O
are	O
more	O
appropriate	O
.	O
</s>
<s>
The	O
actual	O
data	B-Application
mining	I-Application
task	O
is	O
the	O
semi-automatic	O
or	O
automatic	O
analysis	O
of	O
large	O
quantities	O
of	O
data	O
to	O
extract	O
previously	O
unknown	O
,	O
interesting	O
patterns	O
such	O
as	O
groups	O
of	O
data	O
records	O
(	O
cluster	B-Algorithm
analysis	I-Algorithm
)	O
,	O
unusual	O
records	O
(	O
anomaly	B-Algorithm
detection	I-Algorithm
)	O
,	O
and	O
dependencies	B-Application
(	O
association	B-Algorithm
rule	I-Algorithm
mining	I-Algorithm
,	O
sequential	B-Algorithm
pattern	I-Algorithm
mining	I-Algorithm
)	O
.	O
</s>
<s>
These	O
patterns	O
can	O
then	O
be	O
seen	O
as	O
a	O
kind	O
of	O
summary	O
of	O
the	O
input	O
data	O
,	O
and	O
may	O
be	O
used	O
in	O
further	O
analysis	O
or	O
,	O
for	O
example	O
,	O
in	O
machine	O
learning	O
and	O
predictive	B-General_Concept
analytics	I-General_Concept
.	O
</s>
<s>
For	O
example	O
,	O
the	O
data	B-Application
mining	I-Application
step	O
might	O
identify	O
multiple	O
groups	O
in	O
the	O
data	O
,	O
which	O
can	O
then	O
be	O
used	O
to	O
obtain	O
more	O
accurate	O
prediction	O
results	O
by	O
a	O
decision	B-Application
support	I-Application
system	I-Application
.	O
</s>
<s>
Neither	O
the	O
data	O
collection	O
,	O
data	O
preparation	O
,	O
nor	O
result	O
interpretation	O
and	O
reporting	O
is	O
part	O
of	O
the	O
data	B-Application
mining	I-Application
step	O
,	O
although	O
they	O
do	O
belong	O
to	O
the	O
overall	O
KDD	O
process	O
as	O
additional	O
steps	O
.	O
</s>
<s>
The	O
difference	O
between	O
data	B-General_Concept
analysis	I-General_Concept
and	O
data	B-Application
mining	I-Application
is	O
that	O
data	B-General_Concept
analysis	I-General_Concept
is	O
used	O
to	O
test	O
models	O
and	O
hypotheses	O
on	O
the	O
dataset	B-General_Concept
,	O
e.g.	O
,	O
analyzing	O
the	O
effectiveness	O
of	O
a	O
marketing	O
campaign	O
,	O
regardless	O
of	O
the	O
amount	O
of	O
data	O
.	O
</s>
<s>
In	O
contrast	O
,	O
data	B-Application
mining	I-Application
uses	O
machine	O
learning	O
and	O
statistical	O
models	O
to	O
uncover	O
clandestine	O
or	O
hidden	O
patterns	O
in	O
a	O
large	O
volume	O
of	O
data	O
.	O
</s>
<s>
The	O
related	O
terms	O
data	B-Algorithm
dredging	I-Algorithm
,	O
data	B-Algorithm
fishing	I-Algorithm
,	O
and	O
data	B-Algorithm
snooping	I-Algorithm
refer	O
to	O
the	O
use	O
of	O
data	B-Application
mining	I-Application
methods	O
to	O
sample	O
parts	O
of	O
a	O
larger	O
population	O
data	B-General_Concept
set	I-General_Concept
that	O
are	O
(	O
or	O
may	O
be	O
)	O
too	O
small	O
for	O
reliable	O
statistical	O
inferences	O
to	O
be	O
made	O
about	O
the	O
validity	O
of	O
any	O
patterns	O
discovered	O
.	O
</s>
<s>
In	O
the	O
1960s	O
,	O
statisticians	O
and	O
economists	O
used	O
terms	O
like	O
data	B-Algorithm
fishing	I-Algorithm
or	O
data	B-Algorithm
dredging	I-Algorithm
to	O
refer	O
to	O
what	O
they	O
considered	O
the	O
bad	O
practice	O
of	O
analyzing	O
data	O
without	O
an	O
a-priori	O
hypothesis	O
.	O
</s>
<s>
The	O
term	O
"	O
data	B-Application
mining	I-Application
"	O
was	O
used	O
in	O
a	O
similarly	O
critical	O
way	O
by	O
economist	O
Michael	O
Lovell	O
in	O
an	O
article	O
published	O
in	O
the	O
Review	O
of	O
Economic	O
Studies	O
in	O
1983	O
.	O
</s>
<s>
The	O
term	O
data	B-Application
mining	I-Application
appeared	O
around	O
1990	O
in	O
the	O
database	O
community	O
,	O
with	O
generally	O
positive	O
connotations	O
.	O
</s>
<s>
For	O
a	O
short	O
time	O
in	O
1980s	O
,	O
a	O
phrase	O
"	O
database	O
mining	O
"	O
™	O
,	O
was	O
used	O
,	O
but	O
since	O
it	O
was	O
trademarked	O
by	O
HNC	O
,	O
a	O
San	O
Diego-based	O
company	O
,	O
to	O
pitch	O
their	O
Database	O
Mining	O
Workstation	O
;	O
researchers	O
consequently	O
turned	O
to	O
data	B-Application
mining	I-Application
.	O
</s>
<s>
Other	O
terms	O
used	O
include	O
data	O
archaeology	O
,	O
information	O
harvesting	O
,	O
information	O
discovery	O
,	O
knowledge	O
extraction	B-General_Concept
,	O
etc	O
.	O
</s>
<s>
Gregory	O
Piatetsky-Shapiro	O
coined	O
the	O
term	O
"	O
knowledge	B-Application
discovery	I-Application
in	I-Application
databases	I-Application
"	O
for	O
the	O
first	O
workshop	O
on	O
the	O
same	O
topic	O
and	O
this	O
term	O
became	O
more	O
popular	O
in	O
AI	B-Application
and	O
machine	O
learning	O
community	O
.	O
</s>
<s>
However	O
,	O
the	O
term	O
data	B-Application
mining	I-Application
became	O
more	O
popular	O
in	O
the	O
business	O
and	O
press	O
communities	O
.	O
</s>
<s>
Currently	O
,	O
the	O
terms	O
data	B-Application
mining	I-Application
and	O
knowledge	O
discovery	O
are	O
used	O
interchangeably	O
.	O
</s>
<s>
The	O
manual	O
extraction	B-General_Concept
of	O
patterns	O
from	O
data	O
has	O
occurred	O
for	O
centuries	O
.	O
</s>
<s>
As	O
data	B-General_Concept
sets	I-General_Concept
have	O
grown	O
in	O
size	O
and	O
complexity	O
,	O
direct	O
"	O
hands-on	O
"	O
data	B-General_Concept
analysis	I-General_Concept
has	O
increasingly	O
been	O
augmented	O
with	O
indirect	O
,	O
automated	O
data	O
processing	O
,	O
aided	O
by	O
other	O
discoveries	O
in	O
computer	B-General_Concept
science	I-General_Concept
,	O
specially	O
in	O
the	O
field	O
of	O
machine	O
learning	O
,	O
such	O
as	O
neural	B-Architecture
networks	I-Architecture
,	O
cluster	B-Algorithm
analysis	I-Algorithm
,	O
genetic	B-Algorithm
algorithms	I-Algorithm
(	O
1950s	O
)	O
,	O
decision	B-Algorithm
trees	I-Algorithm
and	O
decision	O
rules	O
(	O
1960s	O
)	O
,	O
and	O
support	B-Algorithm
vector	I-Algorithm
machines	I-Algorithm
(	O
1990s	O
)	O
.	O
</s>
<s>
Data	B-Application
mining	I-Application
is	O
the	O
process	O
of	O
applying	O
these	O
methods	O
with	O
the	O
intention	O
of	O
uncovering	O
hidden	O
patterns	O
.	O
</s>
<s>
in	O
large	O
data	B-General_Concept
sets	I-General_Concept
.	O
</s>
<s>
It	O
bridges	O
the	O
gap	O
from	O
applied	O
statistics	O
and	O
artificial	B-Application
intelligence	I-Application
(	O
which	O
usually	O
provide	O
the	O
mathematical	O
background	O
)	O
to	O
database	B-Application
management	I-Application
by	O
exploiting	O
the	O
way	O
data	O
is	O
stored	O
and	O
indexed	O
in	O
databases	B-Application
to	O
execute	O
the	O
actual	O
learning	O
and	O
discovery	O
algorithms	O
more	O
efficiently	O
,	O
allowing	O
such	O
methods	O
to	O
be	O
applied	O
to	O
ever-larger	O
data	B-General_Concept
sets	I-General_Concept
.	O
</s>
<s>
The	O
knowledge	B-Application
discovery	I-Application
in	I-Application
databases	I-Application
(	O
KDD	O
)	O
process	O
is	O
commonly	O
defined	O
with	O
the	O
stages	O
:	O
</s>
<s>
It	O
exists	O
,	O
however	O
,	O
in	O
many	O
variations	O
on	O
this	O
theme	O
,	O
such	O
as	O
the	O
Cross-industry	B-Algorithm
standard	I-Algorithm
process	I-Algorithm
for	I-Algorithm
data	I-Algorithm
mining	I-Algorithm
(	O
CRISP-DM	B-Algorithm
)	O
which	O
defines	O
six	O
phases	O
:	O
</s>
<s>
or	O
a	O
simplified	O
process	O
such	O
as	O
(	O
1	O
)	O
Pre-processing	O
,	O
(	O
2	O
)	O
Data	B-Application
Mining	I-Application
,	O
and	O
(	O
3	O
)	O
Results	O
Validation	O
.	O
</s>
<s>
Polls	O
conducted	O
in	O
2002	O
,	O
2004	O
,	O
2007	O
and	O
2014	O
show	O
that	O
the	O
CRISP-DM	B-Algorithm
methodology	O
is	O
the	O
leading	O
methodology	O
used	O
by	O
data	B-Application
miners	I-Application
.	O
</s>
<s>
The	O
only	O
other	O
data	B-Application
mining	I-Application
standard	O
named	O
in	O
these	O
polls	O
was	O
SEMMA	B-Algorithm
.	O
</s>
<s>
However	O
,	O
3	O
–	O
4	O
times	O
as	O
many	O
people	O
reported	O
using	O
CRISP-DM	B-Algorithm
.	O
</s>
<s>
Several	O
teams	O
of	O
researchers	O
have	O
published	O
reviews	O
of	O
data	B-Application
mining	I-Application
process	O
models	O
,	O
and	O
Azevedo	O
and	O
Santos	O
conducted	O
a	O
comparison	O
of	O
CRISP-DM	B-Algorithm
and	O
SEMMA	B-Algorithm
in	O
2008	O
.	O
</s>
<s>
Before	O
data	B-Application
mining	I-Application
algorithms	O
can	O
be	O
used	O
,	O
a	O
target	O
data	B-General_Concept
set	I-General_Concept
must	O
be	O
assembled	O
.	O
</s>
<s>
As	O
data	B-Application
mining	I-Application
can	O
only	O
uncover	O
patterns	O
actually	O
present	O
in	O
the	O
data	O
,	O
the	O
target	O
data	B-General_Concept
set	I-General_Concept
must	O
be	O
large	O
enough	O
to	O
contain	O
these	O
patterns	O
while	O
remaining	O
concise	O
enough	O
to	O
be	O
mined	O
within	O
an	O
acceptable	O
time	O
limit	O
.	O
</s>
<s>
A	O
common	O
source	O
for	O
data	O
is	O
a	O
data	B-General_Concept
mart	I-General_Concept
or	O
data	O
warehouse	O
.	O
</s>
<s>
Pre-processing	O
is	O
essential	O
to	O
analyze	O
the	O
multivariate	B-General_Concept
data	I-General_Concept
sets	O
before	O
data	B-Application
mining	I-Application
.	O
</s>
<s>
Data	O
cleaning	O
removes	O
the	O
observations	O
containing	O
noise	B-General_Concept
and	O
those	O
with	O
missing	O
data	O
.	O
</s>
<s>
Data	B-Application
mining	I-Application
involves	O
six	O
common	O
classes	O
of	O
tasks	O
:	O
</s>
<s>
Anomaly	B-Algorithm
detection	I-Algorithm
(	O
outlier/change/deviation	O
detection	O
)	O
–	O
The	O
identification	O
of	O
unusual	O
data	O
records	O
,	O
that	O
might	O
be	O
interesting	O
or	O
data	O
errors	O
that	O
require	O
further	O
investigation	O
.	O
</s>
<s>
Association	B-Algorithm
rule	I-Algorithm
learning	I-Algorithm
(	O
dependency	B-Application
modeling	O
)	O
–	O
Searches	O
for	O
relationships	O
between	O
variables	O
.	O
</s>
<s>
Using	O
association	B-Algorithm
rule	I-Algorithm
learning	I-Algorithm
,	O
the	O
supermarket	O
can	O
determine	O
which	O
products	O
are	O
frequently	O
bought	O
together	O
and	O
use	O
this	O
information	O
for	O
marketing	O
purposes	O
.	O
</s>
<s>
Clustering	B-Algorithm
–	O
is	O
the	O
task	O
of	O
discovering	O
groups	O
and	O
structures	O
in	O
the	O
data	O
that	O
are	O
in	O
some	O
way	O
or	O
another	O
"	O
similar	O
"	O
,	O
without	O
using	O
known	O
structures	O
in	O
the	O
data	O
.	O
</s>
<s>
Classification	B-General_Concept
–	O
is	O
the	O
task	O
of	O
generalizing	O
known	O
structure	O
to	O
apply	O
to	O
new	O
data	O
.	O
</s>
<s>
Regression	O
–	O
attempts	O
to	O
find	O
a	O
function	O
that	O
models	O
the	O
data	O
with	O
the	O
least	O
error	O
that	O
is	O
,	O
for	O
estimating	O
the	O
relationships	O
among	O
data	O
or	O
datasets	B-General_Concept
.	O
</s>
<s>
Summarization	B-Application
–	O
providing	O
a	O
more	O
compact	O
representation	O
of	O
the	O
data	B-General_Concept
set	I-General_Concept
,	O
including	O
visualization	B-Application
and	O
report	O
generation	O
.	O
</s>
<s>
Data	B-Application
mining	I-Application
can	O
unintentionally	O
be	O
misused	O
,	O
producing	O
results	O
that	O
appear	O
to	O
be	O
significant	O
but	O
which	O
do	O
not	O
actually	O
predict	O
future	O
behavior	O
and	O
cannot	O
be	O
reproduced	O
on	O
a	O
new	O
sample	O
of	O
data	O
,	O
therefore	O
bearing	O
little	O
use	O
.	O
</s>
<s>
A	O
simple	O
version	O
of	O
this	O
problem	O
in	O
machine	O
learning	O
is	O
known	O
as	O
overfitting	B-Error_Name
,	O
but	O
the	O
same	O
problem	O
can	O
arise	O
at	O
different	O
phases	O
of	O
the	O
process	O
and	O
thus	O
a	O
train/test	O
split	O
—	O
when	O
applicable	O
at	O
all	O
—	O
may	O
not	O
be	O
sufficient	O
to	O
prevent	O
this	O
from	O
happening	O
.	O
</s>
<s>
The	O
final	O
step	O
of	O
knowledge	O
discovery	O
from	O
data	O
is	O
to	O
verify	O
that	O
the	O
patterns	O
produced	O
by	O
the	O
data	B-Application
mining	I-Application
algorithms	O
occur	O
in	O
the	O
wider	O
data	B-General_Concept
set	I-General_Concept
.	O
</s>
<s>
It	O
is	O
common	O
for	O
data	B-Application
mining	I-Application
algorithms	O
to	O
find	O
patterns	O
in	O
the	O
training	O
set	O
which	O
are	O
not	O
present	O
in	O
the	O
general	O
data	B-General_Concept
set	I-General_Concept
.	O
</s>
<s>
This	O
is	O
called	O
overfitting	B-Error_Name
.	O
</s>
<s>
To	O
overcome	O
this	O
,	O
the	O
evaluation	O
uses	O
a	O
test	O
set	B-General_Concept
of	I-General_Concept
data	I-General_Concept
on	O
which	O
the	O
data	B-Application
mining	I-Application
algorithm	O
was	O
not	O
trained	O
.	O
</s>
<s>
For	O
example	O
,	O
a	O
data	B-Application
mining	I-Application
algorithm	O
trying	O
to	O
distinguish	O
"	O
spam	O
"	O
from	O
"	O
legitimate	O
"	O
e-mails	O
would	O
be	O
trained	O
on	O
a	O
training	O
set	O
of	O
sample	O
e-mails	O
.	O
</s>
<s>
Several	O
statistical	O
methods	O
may	O
be	O
used	O
to	O
evaluate	O
the	O
algorithm	O
,	O
such	O
as	O
ROC	B-Algorithm
curves	I-Algorithm
.	O
</s>
<s>
If	O
the	O
learned	O
patterns	O
do	O
not	O
meet	O
the	O
desired	O
standards	O
,	O
it	O
is	O
necessary	O
to	O
re-evaluate	O
and	O
change	O
the	O
pre-processing	O
and	O
data	B-Application
mining	I-Application
steps	O
.	O
</s>
<s>
The	O
premier	O
professional	O
body	O
in	O
the	O
field	O
is	O
the	O
Association	O
for	O
Computing	O
Machinery	O
's	O
(	O
ACM	O
)	O
Special	O
Interest	O
Group	O
(	O
SIG	O
)	O
on	O
Knowledge	O
Discovery	O
and	O
Data	B-Application
Mining	I-Application
(	O
SIGKDD	B-Operating_System
)	O
.	O
</s>
<s>
Since	O
1989	O
,	O
this	O
ACM	O
SIG	O
has	O
hosted	O
an	O
annual	O
international	O
conference	O
and	O
published	O
its	O
proceedings	O
,	O
and	O
since	O
1999	O
it	O
has	O
published	O
a	O
biannual	O
academic	O
journal	O
titled	O
"	O
SIGKDD	B-Operating_System
Explorations	O
"	O
.	O
</s>
<s>
Computer	B-General_Concept
science	I-General_Concept
conferences	O
on	O
data	B-Application
mining	I-Application
include	O
:	O
</s>
<s>
Data	B-Application
mining	I-Application
topics	O
are	O
also	O
present	O
in	O
many	O
data	O
management/database	O
conferences	O
such	O
as	O
the	O
ICDE	O
Conference	O
,	O
SIGMOD	O
Conference	O
and	O
International	O
Conference	O
on	O
Very	O
Large	O
Data	B-Application
Bases	I-Application
.	O
</s>
<s>
There	O
have	O
been	O
some	O
efforts	O
to	O
define	O
standards	O
for	O
the	O
data	B-Application
mining	I-Application
process	O
,	O
for	O
example	O
,	O
the	O
1999	O
European	O
Cross	B-Algorithm
Industry	I-Algorithm
Standard	I-Algorithm
Process	I-Algorithm
for	I-Algorithm
Data	I-Algorithm
Mining	I-Algorithm
(	O
CRISP-DM	B-Algorithm
1.0	O
)	O
and	O
the	O
2004	O
Java	B-Language
Data	I-Language
Mining	I-Language
standard	O
(	O
JDM	O
1.0	O
)	O
.	O
</s>
<s>
Development	O
on	O
successors	O
to	O
these	O
processes	O
(	O
CRISP-DM	B-Algorithm
2.0	O
and	O
JDM	O
2.0	O
)	O
was	O
active	O
in	O
2006	O
but	O
has	O
stalled	O
since	O
.	O
</s>
<s>
For	O
exchanging	O
the	O
extracted	O
models	O
—	O
in	O
particular	O
for	O
use	O
in	O
predictive	B-General_Concept
analytics	I-General_Concept
—	O
the	O
key	O
standard	O
is	O
the	O
Predictive	B-Application
Model	I-Application
Markup	I-Application
Language	I-Application
(	O
PMML	B-Application
)	O
,	O
which	O
is	O
an	O
XML-based	O
language	O
developed	O
by	O
the	O
Data	B-Application
Mining	I-Application
Group	O
(	O
DMG	O
)	O
and	O
supported	O
as	O
exchange	O
format	O
by	O
many	O
data	B-Application
mining	I-Application
applications	O
.	O
</s>
<s>
As	O
the	O
name	O
suggests	O
,	O
it	O
only	O
covers	O
prediction	O
models	O
,	O
a	O
particular	O
data	B-Application
mining	I-Application
task	O
of	O
high	O
importance	O
to	O
business	O
applications	O
.	O
</s>
<s>
However	O
,	O
extensions	O
to	O
cover	O
(	O
for	O
example	O
)	O
subspace	O
clustering	B-Algorithm
have	O
been	O
proposed	O
independently	O
of	O
the	O
DMG	O
.	O
</s>
<s>
Data	B-Application
mining	I-Application
is	O
used	O
wherever	O
there	O
is	O
digital	O
data	O
available	O
today	O
.	O
</s>
<s>
Notable	O
examples	B-Algorithm
of	I-Algorithm
data	I-Algorithm
mining	I-Algorithm
can	O
be	O
found	O
throughout	O
business	O
,	O
medicine	O
,	O
science	O
,	O
and	O
surveillance	O
.	O
</s>
<s>
While	O
the	O
term	O
"	O
data	B-Application
mining	I-Application
"	O
itself	O
may	O
have	O
no	O
ethical	O
implications	O
,	O
it	O
is	O
often	O
associated	O
with	O
the	O
mining	O
of	O
information	O
in	O
relation	O
to	O
user	B-Application
behavior	I-Application
(	O
ethical	O
and	O
otherwise	O
)	O
.	O
</s>
<s>
The	O
ways	O
in	O
which	O
data	B-Application
mining	I-Application
can	O
be	O
used	O
can	O
in	O
some	O
cases	O
and	O
contexts	O
raise	O
questions	O
regarding	O
privacy	O
,	O
legality	O
,	O
and	O
ethics	O
.	O
</s>
<s>
In	O
particular	O
,	O
data	B-Application
mining	I-Application
government	O
or	O
commercial	O
data	B-General_Concept
sets	I-General_Concept
for	O
national	O
security	O
or	O
law	O
enforcement	O
purposes	O
,	O
such	O
as	O
in	O
the	O
Total	O
Information	O
Awareness	O
Program	O
or	O
in	O
ADVISE	O
,	O
has	O
raised	O
privacy	O
concerns	O
.	O
</s>
<s>
Data	B-Application
mining	I-Application
requires	O
data	O
preparation	O
which	O
uncovers	O
information	O
or	O
patterns	O
which	O
compromise	O
confidentiality	O
and	O
privacy	O
obligations	O
.	O
</s>
<s>
This	O
is	O
not	O
data	B-Application
mining	I-Application
per	O
se	O
,	O
but	O
a	O
result	O
of	O
the	O
preparation	O
of	O
data	O
before	O
—	O
and	O
for	O
the	O
purposes	O
of	O
—	O
the	O
analysis	O
.	O
</s>
<s>
The	O
threat	O
to	O
an	O
individual	O
's	O
privacy	O
comes	O
into	O
play	O
when	O
the	O
data	O
,	O
once	O
compiled	O
,	O
cause	O
the	O
data	B-Application
miner	I-Application
,	O
or	O
anyone	O
who	O
has	O
access	O
to	O
the	O
newly	O
compiled	O
data	B-General_Concept
set	I-General_Concept
,	O
to	O
be	O
able	O
to	O
identify	O
specific	O
individuals	O
,	O
especially	O
when	O
the	O
data	O
were	O
originally	O
anonymous	O
.	O
</s>
<s>
The	O
purpose	O
of	O
the	O
data	O
collection	O
and	O
any	O
(	O
known	O
)	O
data	B-Application
mining	I-Application
projects	O
.	O
</s>
<s>
However	O
,	O
even	O
"	O
anonymized	O
"	O
data	B-General_Concept
sets	I-General_Concept
can	O
potentially	O
contain	O
enough	O
information	O
to	O
allow	O
identification	O
of	O
individuals	O
,	O
as	O
occurred	O
when	O
journalists	O
were	O
able	O
to	O
find	O
several	O
individuals	O
based	O
on	O
a	O
set	O
of	O
search	O
histories	O
that	O
were	O
inadvertently	O
released	O
by	O
AOL	O
.	O
</s>
<s>
In	O
the	O
United	O
Kingdom	O
in	O
particular	O
there	O
have	O
been	O
cases	O
of	O
corporations	O
using	O
data	B-Application
mining	I-Application
as	O
a	O
way	O
to	O
target	O
certain	O
groups	O
of	O
customers	O
forcing	O
them	O
to	O
pay	O
unfairly	O
high	O
prices	O
.	O
</s>
<s>
The	O
use	O
of	O
data	B-Application
mining	I-Application
by	O
the	O
majority	O
of	O
businesses	O
in	O
the	O
U.S.	O
is	O
not	O
controlled	O
by	O
any	O
legislation	O
.	O
</s>
<s>
Under	O
European	O
copyright	O
and	O
database	O
laws	O
,	O
the	O
mining	O
of	O
in-copyright	O
works	O
(	O
such	O
as	O
by	O
web	B-Application
mining	I-Application
)	O
without	O
the	O
permission	O
of	O
the	O
copyright	O
owner	O
is	O
not	O
legal	O
.	O
</s>
<s>
Where	O
a	O
database	O
is	O
pure	O
data	O
in	O
Europe	O
,	O
it	O
may	O
be	O
that	O
there	O
is	O
no	O
copyright	O
—	O
but	O
database	O
rights	O
may	O
exist	O
,	O
so	O
data	B-Application
mining	I-Application
becomes	O
subject	O
to	O
intellectual	O
property	O
owners	O
 '	O
rights	O
that	O
are	O
protected	O
by	O
the	O
Database	O
Directive	O
.	O
</s>
<s>
The	O
UK	O
was	O
the	O
second	O
country	O
in	O
the	O
world	O
to	O
do	O
so	O
after	O
Japan	O
,	O
which	O
introduced	O
an	O
exception	O
in	O
2009	O
for	O
data	B-Application
mining	I-Application
.	O
</s>
<s>
Since	O
2020	O
also	O
Switzerland	O
has	O
been	O
regulating	O
data	B-Application
mining	I-Application
by	O
allowing	O
it	O
in	O
the	O
research	O
field	O
under	O
certain	O
conditions	O
laid	O
down	O
by	O
art	O
.	O
</s>
<s>
The	O
European	O
Commission	O
facilitated	O
stakeholder	O
discussion	O
on	O
text	B-Algorithm
and	I-Algorithm
data	I-Algorithm
mining	I-Algorithm
in	O
2013	O
,	O
under	O
the	O
title	O
of	O
Licences	O
for	O
Europe	O
.	O
</s>
<s>
For	O
example	O
,	O
as	O
part	O
of	O
the	B-Application
Google	I-Application
Book	O
settlement	O
the	O
presiding	O
judge	O
on	O
the	O
case	O
ruled	O
that	O
Google	B-Application
's	I-Application
digitization	O
project	O
of	O
in-copyright	O
books	O
was	O
lawful	O
,	O
in	O
part	O
because	O
of	O
the	O
transformative	O
uses	O
that	O
the	O
digitization	O
project	O
displayed	O
—	O
one	O
being	O
text	B-Algorithm
and	I-Algorithm
data	I-Algorithm
mining	I-Algorithm
.	O
</s>
<s>
The	O
following	O
applications	O
are	O
available	O
under	O
free/open	O
-source	O
licenses	O
.	O
</s>
<s>
Carrot2	B-Application
:	O
Text	O
and	O
search	O
results	O
clustering	B-Algorithm
framework	O
.	O
</s>
<s>
Chemicalize.org	B-Application
:	O
A	O
chemical	O
structure	O
miner	O
and	O
web	O
search	O
engine	O
.	O
</s>
<s>
ELKI	B-Language
:	O
A	O
university	O
research	O
project	O
with	O
advanced	O
cluster	B-Algorithm
analysis	I-Algorithm
and	O
outlier	B-Algorithm
detection	I-Algorithm
methods	O
written	O
in	O
the	O
Java	B-Language
language	I-Language
.	O
</s>
<s>
GATE	B-Language
:	O
a	O
natural	B-Language
language	I-Language
processing	I-Language
and	O
language	O
engineering	O
tool	O
.	O
</s>
<s>
KNIME	B-Language
:	O
The	O
Konstanz	O
Information	O
Miner	O
,	O
a	O
user-friendly	O
and	O
comprehensive	O
data	B-General_Concept
analytics	I-General_Concept
framework	O
.	O
</s>
<s>
Massive	B-Language
Online	I-Language
Analysis	I-Language
(	O
MOA	B-Language
)	O
:	O
a	O
real-time	O
big	B-Application
data	I-Application
stream	O
mining	O
with	O
concept	O
drift	O
tool	O
in	O
the	O
Java	B-Language
programming	I-Language
language	I-Language
.	O
</s>
<s>
MEPX	B-Algorithm
:	O
cross-platform	O
tool	O
for	O
regression	O
and	O
classification	B-General_Concept
problems	O
based	O
on	O
a	O
Genetic	O
Programming	O
variant	O
.	O
</s>
<s>
mlpack	B-Language
:	O
a	O
collection	O
of	O
ready-to-use	O
machine	O
learning	O
algorithms	O
written	O
in	O
the	O
C++	B-Language
language	I-Language
.	O
</s>
<s>
NLTK	B-General_Concept
(	O
Natural	B-General_Concept
Language	I-General_Concept
Toolkit	I-General_Concept
)	O
:	O
A	O
suite	O
of	O
libraries	O
and	O
programs	O
for	O
symbolic	O
and	O
statistical	O
natural	B-Language
language	I-Language
processing	I-Language
(	O
NLP	B-Language
)	O
for	O
the	O
Python	B-Language
language	I-Language
.	O
</s>
<s>
OpenNN	B-Language
:	O
Open	O
neural	B-Architecture
networks	I-Architecture
library	O
.	O
</s>
<s>
Orange	B-Application
:	O
A	O
component-based	O
data	B-Application
mining	I-Application
and	O
machine	O
learning	O
software	O
suite	O
written	O
in	O
the	O
Python	B-Language
language	I-Language
.	O
</s>
<s>
R	B-Language
:	O
A	O
programming	O
language	O
and	O
software	O
environment	O
for	O
statistical	O
computing	O
,	O
data	B-Application
mining	I-Application
,	O
and	O
graphics	O
.	O
</s>
<s>
It	O
is	O
part	O
of	O
the	B-Application
GNU	I-Application
Project	I-Application
.	O
</s>
<s>
scikit-learn	B-Application
:	O
An	O
open-source	B-License
machine	O
learning	O
library	O
for	O
the	O
Python	B-Language
programming	I-Language
language	I-Language
;	O
</s>
<s>
Torch	B-Algorithm
:	O
An	O
open-source	B-License
deep	B-Algorithm
learning	I-Algorithm
library	O
for	O
the	O
Lua	B-Language
programming	I-Language
language	I-Language
and	O
scientific	O
computing	O
framework	O
with	O
wide	O
support	O
for	O
machine	O
learning	O
algorithms	O
.	O
</s>
<s>
UIMA	B-Application
:	O
The	O
UIMA	B-Application
(	O
Unstructured	O
Information	O
Management	O
Architecture	O
)	O
is	O
a	O
component	O
framework	O
for	O
analyzing	O
unstructured	O
content	O
such	O
as	O
text	O
,	O
audio	O
and	O
video	O
–	O
originally	O
developed	O
by	O
IBM	O
.	O
</s>
<s>
Weka	B-Language
:	O
A	O
suite	O
of	O
machine	O
learning	O
software	O
applications	O
written	O
in	O
the	O
Java	B-Language
programming	I-Language
language	I-Language
.	O
</s>
<s>
LIONsolver	B-Application
:	O
an	O
integrated	O
software	O
application	O
for	O
data	B-Application
mining	I-Application
,	O
business	B-General_Concept
intelligence	I-General_Concept
,	O
and	O
modeling	O
that	O
implements	O
the	O
Learning	O
and	O
Intelligent	O
OptimizatioN	O
(	O
LION	O
)	O
approach	O
.	O
</s>
<s>
PolyAnalyst	B-Application
:	O
data	B-Algorithm
and	I-Algorithm
text	I-Algorithm
mining	I-Algorithm
software	O
by	O
Megaputer	O
Intelligence	O
.	O
</s>
<s>
Microsoft	B-Application
Analysis	I-Application
Services	I-Application
:	O
data	B-Application
mining	I-Application
software	O
provided	O
by	O
Microsoft	O
.	O
</s>
<s>
NetOwl	B-Algorithm
:	O
suite	O
of	O
multilingual	O
text	O
and	O
entity	O
analytics	O
products	O
that	O
enable	O
data	B-Application
mining	I-Application
.	O
</s>
<s>
Oracle	B-Algorithm
Data	I-Algorithm
Mining	I-Algorithm
:	O
data	B-Application
mining	I-Application
software	O
by	O
Oracle	B-Application
Corporation	I-Application
.	O
</s>
<s>
PSeven	B-Algorithm
:	O
platform	O
for	O
automation	O
of	O
engineering	O
simulation	O
and	O
analysis	O
,	O
multidisciplinary	O
optimization	O
and	O
data	B-Application
mining	I-Application
provided	O
by	O
DATADVANCE	B-Algorithm
.	O
</s>
<s>
Qlucore	B-Application
Omics	O
Explorer	O
:	O
data	B-Application
mining	I-Application
software	O
.	O
</s>
<s>
RapidMiner	B-Algorithm
:	O
An	O
environment	O
for	O
machine	O
learning	O
and	O
data	B-Application
mining	I-Application
experiments	O
.	O
</s>
<s>
SAS	O
Enterprise	O
Miner	O
:	O
data	B-Application
mining	I-Application
software	O
provided	O
by	O
the	O
SAS	O
Institute	O
.	O
</s>
<s>
SPSS	B-Algorithm
Modeler	I-Algorithm
:	O
data	B-Application
mining	I-Application
software	O
provided	O
by	O
IBM	O
.	O
</s>
<s>
STATISTICA	B-Application
Data	B-Application
Miner	I-Application
:	O
data	B-Application
mining	I-Application
software	O
provided	O
by	O
StatSoft	O
.	O
</s>
<s>
Tanagra	B-Language
:	O
Visualisation-oriented	O
data	B-Application
mining	I-Application
software	O
,	O
also	O
for	O
teaching	O
.	O
</s>
<s>
Vertica	O
:	O
data	B-Application
mining	I-Application
software	O
provided	O
by	O
Hewlett-Packard	O
.	O
</s>
<s>
Google	B-General_Concept
Cloud	I-General_Concept
Platform	I-General_Concept
:	O
automated	O
custom	O
ML	O
models	O
managed	O
by	O
Google	B-Application
.	I-Application
</s>
<s>
Amazon	B-General_Concept
SageMaker	I-General_Concept
:	O
managed	O
service	O
provided	O
by	O
Amazon	B-Application
for	O
creating	O
&	O
productionising	O
custom	O
ML	O
models	O
.	O
</s>
