Elevate Your Advertising Game: Introducing Billion Broadcaster Lift Advertising
Big Data in Search of Meta Data
1. Big
Data
in
Search
of
Meta
Data
Prof
Dr
Dirk
Baecker
Content
Strategy
Forum
Conference,
Frankfurt
am
Main,
1-‐3
July
2014
2. Metadata
Defined
Metadata
are
not
just
about
• addresses,
• (strong
and
weak)
connecHons,
• (high
and
low)
frequencies,
• structural
holes,
• bridges,
• and
structural
equivalencies
(network
analysis)
but
also
about
any
kind
of
data
able
to
sort
data
(Bagley
1968).
Which
means
that
you
have
to
know
how
to
ask
in
order
to
be
able
to
find
out.
2
3. E.g.,
the
Influence
Matrix
Prob
(ht
(c')
|
ht-‐1
(1),
…,
ht-‐1
(C))
with
h
=
hidden
behavior
of
a
person
c
of
a
group
of
people
C.
3
New
York
2014
4. Or
else,
an
Ac<on
Matrix
Prob
(ht
(c')
|
ht-‐1
(1),
…,
ht-‐1
(C)|
ht+1
(1),
…,
ht+1
(C))
h
=
hidden
behavior
determined
by:
• purpose/system
and
• future/evoluHon,
within
the
limits
of:
• complexity/
environment,
with
t+1
=
unknown,
C
=
fluctuaHng,
and
Prob
=
uncertain.
4
5. A
Metadata
Matrix
• A
Calculus
of
Limits.
• Behavior
is
unpredictable
because
it
is
structure-‐
determined
–
and
you
don't
know
which
structure
prevails.
• Content
strategy
is
about
possible,
yet
improbable
in-‐
form-‐ac4on.
• Pick
your
favorite
observer.
• Ask
for
metadata
that
let
you
correct
your
expectaHons
of
observers'
behavior.
5