Improved letters about benefits A case study

Documents
18 pages
Obtenez un accès à la bibliothèque pour le consulter en ligne
En savoir plus

Description

  • leçon - matière potentielle : style
  • leçon - matière potentielle : strategy
  • fiche de synthèse - matière potentielle : level
  • fiche de synthèse - matière potentielle : table
  • exposé - matière potentielle : reasons
  • expression écrite
The Department for Work & Pensions deal with most of the UK's social security benefits, and send out over 150 million letters a year. Their system-produced letters had become long and unwieldy, so they asked the Simplification Centre to join their project team tasked with designing a better format. The new format breaks up the content, showing its different status, and uses a layout that encourages strategic reading. It is written in plain English, and explains step-by-step how amounts are calculated.
  • dwp
  • line length
  • team around halfway through the process
  • letters about benefits
  • panels
  • document
  • team
  • content

Sujets

Informations

Publié par
Nombre de visites sur la page 18
Langue English
Signaler un problème

Chapter
Overview Queueing
l
configurations. hardware various on running
when provide will system the that times response the project must
vendor this types, various of transactions of rates arrival the of
estimates Given bids. preparing in systems size to way efficient an
needs system information medical turn-key a of manufacturer The -
acquisition and sizing during
packet? single a in multiplexed be frame
main- same the for destined terminals multiple from characters
Should characters? multiple contain packets Should clustered?
be terminals Should network. communications broadcast oriented
packet- a using mainframes to terminals connecting for protocols
and architectures various considering is manufacturer computer A -
implementation.
to prior evaluated be must mechanisms various of merits tive
rela- The mechanism. communication and synchronization process
the and mechanism accessing database the as such issues on made
be must decisions fundamental phase, design the in Early tions.
worksta- graphics through database distributed a to access taneous
simul- designers aircraft hundred several allow to system design
computer-aided a building and designing is company aerospace An -
implementation and design during 0
system: a of life the throughout arise that performance and cost
of questions the to answers intelligent provide to necessary is standing
under- an Such systems. these of behavior the understanding in assist
that techniques and tools for need increasing an is result The ago. years
few a even of those than business of conduct the to essential more and
evolving, rapidly more complex, more are systems computer Today’s
Introduction 1.1.
Modelling Network of An
1 l
modelling. book: this of ject
sub- general the approach, third a lies extremes these Between inflexible.
and laborious is but accuracy, excellent yields Experimentation verify.
and acquire to difficult are that insight and experience on relies it because
suspect is accuracy its but flexible, and rapid is Intuition spectrum.
a of extremes opposite at sense some in are approaches two These
eralization.
gen- allow would that insight any not but assumptions, of set one under
behavior system of knowledge accurate yield to likely is experiment an
that is drawback further A so. prohibitively often - expensive is also It
choice. of approach the sometimes and required, often valuable, always is
Experimentation alternatives. of evaluation experimental the is Another
rare. are quantity sufficient
in qualities these possess who those Unfortunately, intuition. reliable
yields that insight and experience of degree the for substitutes few are
there sure, be To extrapolation. trend and intuition of use the is One
available. are approaches several basis, a as this With study.
the of objectives the and application, the system, the of grasp thorough
a with begin must one these, as such questions considering In
obtained.
easily not are answers correct complex; also are questions these tunately,
Unfor- answers, incorrect from repercussions serious potentially with
involved, organizations the to significance great of are questions These
evaluated. be must speed device paging and power, CPU size,
memory among tradeoffs system, memory virtual a is this since
particular, In it. alleviating for alternatives various of effectiveness
cost- relative the and be, will bottleneck system the what know
to desirable is It growth. workload of forecasts given configuration,
current its of longevity the assess must utility energy An -
implemented. is change the
when place in be must personnel, and computer both resources,
Adequate seven. of factor a by increase to expected is tions
transac- options of volume total exchange’s the occurs, this When
options. of class new a trading begin to intends exchange stock A -
workload and conjiguration the of evolution during
decision.
a reach to which in available is month one Only procurement.
the to essential is systems twenty these of capacity the comparing
requirements, mandatory c~ertain meeting systems those among
port” per “cost the is criterion selection the Since instruction.
undergraduate for computing interactive provide to proposals
for request a to response in bids twenty received has university A -
3 Introduction 1.1. What
model. network queueing degenerate) (somewhat a constitute
customers arriving its and center service this fact, In depart. and server,
the from service receive necessary, if queue the in wait center, service
the at arrive Customers center. service single a illustrates 1.1 Figure
Centers Service Single 1.2.1.
1.5.) Section in noted be will that liberties
certain take follows, that overview informal the and definitions, (These
parameters. its and queues of network the by induced equations of set
a efficiently solve to software using involves evaluation Analytic sactions.
tran- or users represent which customers, and resources, system represent
which centers, service of collection a is queues of network A cally.
analyti- evaluated is which queues of network a as represented is system
computer the which in modelling system computer to approach ticular
par- a is book, this of subject specific the modelling, network Queueing
Model? Network Queueing a Is 1.2.
system. computer a about information understanding and evaluating,
organizing, gathering, for framework a provides then, Modelling,
model. the parameterize
and define to order in necessary are experiments which to as dance
gui- gives modelling to approach particular each by provided framework
the because enhanced is Experimentation measures.) performance its of
values specific necessarily not but system a of behavior general the reflect
accurately which models, qualitative detailed less by provided be can tion
intui- to guide effective equally an system, a of measures performance the
reflect accurately which models, quantitative devise to is book this in tive
objec- our although fact, (In alternatives. of range wide a under system
a of behavior the investigate to - hunches” “pursue to possible it makes
model a because enhanced is Intuition experimentation. and intuition
both enhances model a using importance, equal Of models. of evaluation
and parameterization, definition, the for framework a provides modelling
to approach particular each methodical: more is it because intuition,
than reliable more is It detail. unnecessary avoids that abstraction an is
model the because experimentation, than flexible more and laborious less
is behavior system investigate to model a Using alternative. this under
behavior its determine to evaluated then and study, under alternatives the
of any reflect to parameterized be can it process, abstraction this through
defined been has model a Once behavior. system’s the to essential
are that aspects those exactly itself, system the is that details of mass
the from distill, to attempt an system: a of abstraction an is model A
Modelling Network Queueing of Overview An Preliminaries: 4 the As gradual. is increase this Initially, it. with along time residence
and increases, congestion rises, intensity workload the As requirement.
service its to equal roughly time residence a have will and immediately
service enter will so center, service the for competition encounter
will seldom customer arriving an that expect we low, is intensity load
work- the When time. residence Consider experience. and intuition with
consistent qualitatively are that measures performance yields model the of
evaluation the that is 1.2 Figure about observe to thing principal The
saturated. be will center service the then this, than greater
is rate arrival the if customers/second; 0.8 or seconds, 1.25 every one
is customers handle can center service the which at rate possible greatest
the seconds, 1.25 is customer a of requirement service average the that
given end, high the On zero. than less be to rate arrival the for sense no
makes it end, low the On parameter. this for values of range interesting
the is This arrivals/second. 0.8 to 0.0 from varies intensity workload the
as measures performance these of each graph 1.2b and 1.2a Figures
customers/second 0.5 throughput:
customers 1.67 length: queue
seconds 3.33 time: residence
.625 utilization:
are: measures performance these later) stated be will that assumptions
certain (under values parameter example our For center). service
the through pass customers which at rate (the throughput and service),
receiving and waiting both center, service the at customers of number
average (the length queue service), receiving and queueing both customer,
a by center service the at spent time average (the time residence busy),
is server the time of proportion (the utilization as such measures formance
per- yielding equations, simple some solving by model this evaluate to ble
possi- is it values, parameter specific For seconds). 1.25 (e.g., customer
a of requirement service average the is which demand, service the specify
must we Second, customers/second). 0.5 or seconds, two every customer
one (e.g., arrive customers which at rate the is case this in which intensity,
workload the specify must we First, parameters. two has model This
Center Service Single A - 1.1 Figure
customers customers
Arriving Departing
Server Queue
5 Model? Network Queueing a Is What 1.2. 0.8
Center Service Single the for Measures Performance - 1.2a Figure
Arrivals/second
0.0
o-
Time: Residence
Arrivals/second
Utilization:
Modelling Network Queueing of Overview An Preliminaries: 6 0.0
0.8
0.0
0.0
0.8
0.8
I
l-
P
Center Service Single the for Measures Performance - 1.2b Figure
Arrivals/second
Throughput:
Arrivals/second
Length: Queue
7 ? Model Network Queueing a Is Kkat 1.2. We
We
that specify might we example, For matters.) center each at demand
service total the only however; important, not is centers the among
circulation of pattern The departing. then and centers, service the among
circulating arriving, as customers of think can we figure, the in lines
the by indicated (As system. the in resource corresponding the at saction
tran- per requirement service total the to corresponds center service each
at demand service the and system, the to transactions submit users which
at rate the to corresponds intensity workload the then system, the in tions
transac- to corresponding as model the in customers view we If center.
service each for demand service separate a provide we time this but
demand, service the specify must also arrive. customers which at rate
the is again once which intensity, workload the specify must one.
previous the of those to analogous are model this of parameters The
Queues of Network A - 1.3 Figure
customers
customers
Disks Departiny
center. service separate a by represented
is disks) three and CPU a case this (in resource system each which in
model realistic more a shows 1.3 Figure 1960’s.) the of days simpler the
in times several success with done was this however, fact, (In 1.1. Figure
of model the use to order in required be would as parameters, two by
system computer contemporary a characterizing imagine to hard is It
Centers Service Multiple 1.2.2.
time. residence in increases dramatic in result rate arrival
in increases small saturation, approaches center service the as until,
rate, faster and faster a at increases time residence however, grows, load
Modelling Network Queueing of Overview An Preliminaries: 8 (We
of model the modified have we 1.4 Figure In workloads.) interactive ing
‘represent- to suited well is alternative (This interactions. between minal)
ter- a uses (i.e., thinking” “spends customer each that time average the
and customers of number the specify to is approach third A workloads.)
batch representing to suited well is alternative (This model. the in mers
custo- of number the state to is approach second A intensity. workload
describe to means available three of one only is workloads) processing
transaction certain representing to suited well is that approach (an arrive
customers which at rate the specifying richness, this of example an As
systems. computer with correspondence the extending far, thus illustrated
have we than attributes of set richer a have models network Queueing
systems. computer in transactions or users to
correspond naturally models network queueing in customers and systems,
computer in queues software their and resources hardware to correspond
naturally models network queueing in centers service example, For tems.
sys- computer of attributes the and models network queueing of butes
attri- the between correspondence close the by straightforward relatively
made is system particular a of model network queueing a Defining
Definition 1.3.1.
Models Netiork
Queueing Evaluating and Parameterizing, Defining, 1.3.
omitted.) is it if even stood
under- be should “average” word the Thus times). response of centile
per- 90th the (e.g., information distributional than rather time) response
average (e.g., values average are models network queueing from obtained
measures performance Most time. response system perceived of notion
intuitive the to correspond to time response and customer, a by center vice
ser- a at spent time the mean to time residence use will consistently
transactions/second 0.2 throughput: system
6.4 transactions: active concurrently of number average
seconds 32.1 users: by perceived time response system average
.60 utilization: CPU
include: measures performance. later)
stated be will that assumptions certain (under values parameter example
our For equations. simple some solving by model this evaluate to ble
possi- is it values parameter specific for center, service single the of case
the in As disks. three the at respectively, service, of seconds 4 and 2,
1, and CPU the at service of seconds 3 of average an requires transaction
such each that and seconds, five every one of rate a at arrive transactions
9 Models Network Queueing Evaluating and Parameterizing, DeJining, 1.3. development program interactive 10 and interactions, between minutes
two of average an for thinks whom of each users database interactive
25 batch, background for 2 of level multiprogramming a processing,
transaction for requests/minute 10 of rate arrival an example, (for class
each for intensities workload provide would we model, the parameterizing
In centers. service relevant the and classes customer four specify would
we model, the defining In development. program interactive and inquiry,
database interactive batch, background processing, transaction ponents:
com- workload four are there which in system computer a directly model
to possible is it example, For demands. service and described) have
we ways the of any in (specified intensity workload own its has which
of each classes, customer multiple of use making by model network
,queueing a in components workload system’s a between distinguish to
possible is it behavior), same the essentially exhibit customers (all class
customer single a had have far thus considered have we that models work
net- queueing the although and components, workload identifiable several
have systems computer most richness, this of example another As
Workload Terminal-Driven a with Model A - 1.4 Figure
CPU
Terminals
Disks
ence.
experi- and intuition with consistent qualitatively is model the of behavior
the that note we again Once seconds. 30 of time think average an with
each users, interactive 50 to 1 from of consists workload the when disks)
three the at respectively, service, of seconds 4 and 2, 1, and CPU the
at service of seconds (3 demands service original the with model this for
utilization CPU and time response system graphs 1.5 Figure approach.
last this using described is intensity workload the that so 1.3 Figure
Modelling Network Queueing of Overview An Preliminaries: 10 200
50
25
1
0.0
*
Model Terminal-Driven the for Measures Performance - 1.5 Figure
users Interactive
r
Time: Response System
users Interactive
Utilization: CPU
11 Models Network Queueing Evaluating and Parameterizing, Defining, 1.3.