TAUS Roundtable Moscow, CAT or TMS Implementation-Calculation of the Number of Licenses and the Total Cost of Ownership, Renat Bikmatov, Logrus International
Ähnlich wie TAUS Roundtable Moscow, CAT or TMS Implementation-Calculation of the Number of Licenses and the Total Cost of Ownership, Renat Bikmatov, Logrus International
Ähnlich wie TAUS Roundtable Moscow, CAT or TMS Implementation-Calculation of the Number of Licenses and the Total Cost of Ownership, Renat Bikmatov, Logrus International (20)
TAUS Roundtable Moscow, CAT or TMS Implementation-Calculation of the Number of Licenses and the Total Cost of Ownership, Renat Bikmatov, Logrus International
1. THURSDAY,
22
May
/14:30
–
15:00
CAT
or
TMS
ImplementaAon:
CalculaAon
of
the
Number
of
Licenses
and
the
Total
Cost
of
Ownership
Renat
Bikmatov,
Logrus
Interna4onal
TAUS
ROUNDTABLE
2014
22
May/
Moscow
(Russia)
2. TranslaAon
AutomaAon
Tools:
The
Use
Case
Big
Data
AnalyAcs
in
LocalizaAon
Industry
The
Presenta+on
Topics:
• How
LSPs
could
minimize
the
cost
of
ownership
of
the
tools
• Live
demo:
the
calculaAon
of
number
of
CAT
licenses
• Which
Big
Data
and
business
intelligence
features
LSPs
might
need
3. TranslaAon
AutomaAon
Tools:
The
Cost
Factors
How
LSPs
Could
Minimize
the
Cost
of
Ownership
of
the
Tools
• Different
types
of
licenses
for
different
workflow
roles:
o project
manager
o linguist
(translator,
editor,
proofreader)
• Flexible
licensing
schemas:
o fixed
license
bound
to
PC
or
user
without
expiraAon
date
o mobile
license
transferable
between
users
o temporary
license
purchased
on
demand
for
limited
period
of
Ame
o license
for
online
connecAon
to
the
translaAon
server
o etc.
• OpportuniAes
for
LSP:
o adjusAng
the
configuraAon
of
the
pool
of
licenses
o minimizing
the
total
cost
of
licenses
4. TranslaAon
AutomaAon
Tools:
The
Live
Demo
CalculaAon
of
Number
of
CAT
Licenses
• The
Task:
to
minimize
the
annual
cost
of
ownership
of
translator
licenses
• The
Method:
to
play
around
with
modifying
parameters
of
licensing
schema
and
calculate
the
number
of
translator
licenses
needed
and
their
annual
cost
in
each
case
• The
Source
Data:
daily
staAsAcs
of
translaAon
tasks
handed
off
to
translators
and
delivered
back
for
all
translaAon
projects
got
from
selected
client(s)
during
the
past
year(s)
6. TranslaAon
AutomaAon
Tools:
The
Live
Demo
The
Summary
of
the
Live
Demo
The
conclusions:
• The
rental
longer
than
30
days
is
not
cost
saving
(the
opAmal
term
of
lease
would
be
5
days
with
30%
savings
in
total
cost)
• The
minimal
number
of
licenses
in
the
rental
package
(5)
is
unessenAal
factor
7. TranslaAon
AutomaAon
Tools:
More
Use
Cases
Big
Data
and
Business
Intelligence
That
LSPs
Might
Need
Big
Data
and
Business
Intelligence
(BI)
clear
the
way
to
further
enhancements
in
LSP’s
business.
Let’s
consider
just
a
few
possible
benefits
using
the
following
template:
• The
translaAon
project/task
parameters
to
be
used
as
input
data
for
BI
tools
• The
BI
reports
generated
from
the
source
data
using
business
analyAc
methods
• The
business
decisions
that
can
be
made
using
the
generated
BI
reports
8. TranslaAon
AutomaAon
Tools:
More
Use
Cases
(Cont.)
Comparison
of
Different
CAT
or
TMS
Products
Input
data:
• CAT
tool:
The
name
of
CAT
tool
used
by
translators
on
given
translaAon
project
• Project
cost:
The
invoiced
price
of
the
translaAon
project
BI
reports:
• DistribuAon
of
company
revenue
by
CAT
tools
in
use:
o %
of
total
revenue
provided
by
each
CAT
tool
o The
volume
of
revenue
provided
by
each
CAT
tool
The
business
quesAons
to
be
answered:
• Which
of
CAT
tools
in
use
could
be
replaced
by
compeAtors
(if
any)
to
reduce
the
cost
of
the
tool
ownership?
• Which
of
TranslaAon
Memory
Server
products
would
be
the
cheapest
soluAon
to
support
the
given
pool
of
CAT
clients
already
in
use?
9. TranslaAon
AutomaAon
Tools:
More
Use
Cases
(Cont.)
UAlizaAon
of
Available
Pool
of
Licenses
Input
data:
• CAT
license
ID:
The
unique
idenAfier
of
each
CAT
license
(if
supported
by
the
license
server)
BI
reports:
• DistribuAon
of
uAlizaAon
percentages
by
CAT
license
IDs
The
business
quesAons
to
be
answered:
• For
how
many
licenses
is
the
uAlizaAon
below
a
predicted
threshold
value?
• Does
it
make
sense
to
adjust
the
licensing
schema
secngs
so
as
to
fit
it
to
the
usage
scenario?
10. TranslaAon
AutomaAon
Tools:
More
Use
Cases
(Cont.)
IdenAfying
Bodlenecks
in
LocalizaAon
Workflow
Input
data:
• Translator’s/Editor’s
ProducAvity
Measurement
Log
records
BI
reports:
• DistribuAon
of
Ame
spent
by
the
translator
or
editor
on
different
subtasks:
translaAon
of
content;
searching
glossaries
for
term
translaAons;
reading
style
guides;
reading
other
reference
materials
provided
for
parAcular
project;
searching
Internet
for
addiAonal
reference
informaAon;
filling
in
feedback
or
Issue
Tracking
forms;
etc.
The
business
quesAons
to
be
answered:
• Which
acAviAes
other
than
translaAng
or
ediAng
as
primary
job
take
most
of
translator’s/editor’s
Ame?
• How
much
Ame
is
spent
on
supporAng
acAviAes
if
compared
to
translaAng
or
ediAng?
• Are
there
any
supporAng
acAviAes
that
might
be
further/beder
automated
to
save
more
Ame?
11. TranslaAon
AutomaAon
Tools:
More
Use
Cases
(Cont.)
Measuring
Machine
TranslaAon
Quality
and
Post-‐editor’s
ProducAvity
Input
data:
• Translator’s/Editor’s
ProducAvity
Measurement
Log
records
BI
reports:
• DistribuAon
of
Ame
spent
by
the
post-‐editor
on
different
TranslaAon
Units
containing
machine
translaAons
The
business
quesAons
to
be
answered:
• What
is
the
average
personal
producAvity
of
each
of
post-‐editors?
This
informaAon
needed
for
planning
resources
for
translaAon
projects
and
calculaAng
personal
rates.
• Which
pieces
of
MT
output
are
worse
and
which
are
the
best?
The
feedback
on
both
data
categories
would
help
to
enhance
the
MT
engine.
• The
average
post-‐ediAng
Ame
normalized
by
personal
producAvity
could
be
used
as
indirect
measure
of
MT
quality.
• If
any
enhancements
have
been
implemented
in
localizaAon
workflow,
is
there
any
impact
on
the
overall
translator’s
or
editor’s
producAvity?
12. TranslaAon
AutomaAon
Tools:
More
Use
Cases
(Cont.)
Measuring
Translator’s
Skill
Level
and
Domain
of
ExperAse
Input
data:
• Translator’s/Editor’s/Proofreader’s
ID
or
Name
• Reviewer’s
feedback
report
as
a
set
of
quality
metrics
of
output
content
BI
reports:
• DistribuAon
of
quality
raAng
of
given
translator
by
the
domains
of
translated
content
The
business
quesAons
to
be
answered:
• In
which
domains
the
translator
produce
translaAons
with
a
quality
score
of
“good”
or
above?
o This
informaAon
would
allow
to
implement
generally
accepted
translator’s
cerAficate
of
quality
and
domain
of
experAse.
o The
list
of
domains
of
experAse
can
be
used
for
automated
selecAon
of
translators
for
incoming
translaAon
projects
in
TranslaAon
Management
System.
o The
list
of
domains
of
insufficient
experAse
can
be
used
as
a
guide
for
translator’s
self-‐educaAon
13. TranslaAon
AutomaAon
Tools:
More
Use
Cases
(Cont.)
UAlizaAon
of
Pool
of
Freelance
Translators
and
Editors
Input
data:
• Translator’s/Editor’s/Proofreader’s
ID
or
Name
BI
reports:
• %
of
uAlizaAon
of
given
translator/editor
in
translaAon
projects
within
his/her
domain
of
experAse
The
business
quesAons
to
be
answered:
• Are
there
any
freelancers
who
are
underused?
o Underused
freelancers
usually
tend
to
cease
cooperaAon
with
the
employer.
As
a
result,
LSP
might
loose
the
money
spent
on
selecAon
and
training
of
the
translators
and
have
to
spend
more
to
replace
the
leakage
of
employees
14. TranslaAon
AutomaAon
Tools:
More
Use
Cases
(Cont.)
Automated
Control
of
Quality
of
Service:
NDA
Agreements
Input
data:
• Client’s
instrucAon
not
to
transfer
the
content
to
any
Cloud
locaAons
(TMS,
file
sharing
services,
etc.)
-‐
OR
-‐
• Client’s
instrucAon
not
to
transfer
the
content
to
specific
online
machine
translaAon
service
BI
reports:
• An
alert
to
be
issued
in
case
of
improper
configuraAon
of
any
translaAon
project
received
from
that
client
The
business
quesAons
to
be
answered:
• Are
the
requirements
of
client’s
NDA
met
in
each
translaAon
project?
15. TranslaAon
AutomaAon
Tools:
More
Use
Cases
(Cont.)
Automated
Control
of
Quality
of
Service:
ConnecAng
Proper
Style
Guides
and
Glossaries
Input
data:
• Style
Guide
ID
or
Name:
The
client
might
have
different
style
guides
and
glossaries
for
different
content
domains
BI
reports:
• An
alert
to
be
issued
in
case
of
improper
configuraAon
of
any
translaAon
project
received
from
that
client.
Typically,
clients
do
not
duplicate
all
the
instrucAon
in
each
handoff,
so
the
project
manager
at
LSP
should
keep
an
eye
on
projects
secngs.
The
domain
of
the
handoff
might
be
specified
in
email
or
encrypted
in
project
package
name.
The
business
quesAons
to
be
answered:
• Are
the
requirements
of
client’s
translaAon
instrucAons
met
in
each
translaAon
project?
16. TranslaAon
AutomaAon
Tools
CalculaAon
of
the
Number
of
Licenses,
EvaluaAon
of
the
Total
Cost
of
Ownership,
and
Other
ApplicaAons
of
Business
Intelligence
THANK
YOU!
QuesAons?