CHUNGWEI HANG MUNINDAR P Singh A Moini Content Serviceoriented computing Preview papers key idea Probabilistic service selection amp composition approaches Experimental results ID: 427309
Download Presentation The PPT/PDF document "Trustworthy Service Selection and Compos..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Trustworthy Service Selection and Composition
CHUNG-WEI HANG MUNINDAR P. Singh
A.
MoiniSlide2
Content
Service-oriented computing
Preview (paper’s key idea)
Probabilistic service selection & composition
approaches
Experimental results
Summary Slide3
Service-Oriented Computing
Every computing resource is packaged as a service
Services are application
building
blocks
unit
of functionality
unit
of integration
unit of composition
Individual services can be “composed” to create more “composite” services
Service have dependencies on other constituent services
C
onsumer service does not have any knowledge of dependencies of services it consumesSlide4
Service-Oriented Computing
Challenges
Services composition (binding) is a design time activity
based on
functional properties
meeting consumer requirements, not
quality attributes
functional properties:
service
types, published as WSDL contract
quality attributes
: throughput, response time, availability..
Service
quality varies by service instance and over time
service quality may change unpredictably Slide5
Contributions
Probabilistic trusted-aware service selection and composition model
takes into account service consumer’s requirements (e.g. qualities of service)
takes into account service
composition patterns
considers
service qualities
as they
apply to
service instances
quality
component service
may affect
the whole
composition
Example:
reliability
rewards & punishes
constituent
services
dynamicallySlide6
Service Composition Patterns
(BEPL Primitives)
SWITCH
chooses exactly one component based on some criteria
MAX
composes quality by inheriting from child with highest quality value
MIN
composes quality by inheriting from
child
with
lowest
quality
throughput for
sequence
SUM
yields composite
quality value as
sum
of
quality
values
obtained from
all
constituent services
PRODUCT
yields
composite
quality value as
product
of
quality
values obtained from all
constituent services.Slide7
http://www.deltalounge.net/wpress/tag/soa-suite/page/2
/
BEPL Services DiagramSlide8
Trust-Aware Service Selection ModelSlide9
Trust-Aware Service
Selection Model
Trustworthiness of
a service
is estimated
based on
direct experience
previous
QoS
received from
service
C
onsumer
maintains its own local model to
determine if to
reward or penalize services based on
direct experience
selects
services
and composes them into a composite
service
evaluates
composite
service with respect
to service quality attributes
applies
a
learning method
to update
its model for
the
services
Special case:
when selecting
atomic service
, consumer
has
less information
to learn fromSlide10
Trust-Aware Service
Selection Models
Two Alternatives
Bayesian Model
m
odels
compositions via
Bayesian networks in partially observable settings
captures
dependency
among
composite and constituent services
adaptively updates
trust to reflect the most recent
quality
uses
online learning to track service behavior and shows how
composite
service’s quality depends upon its constituents’
quality
Beta mixture Model
can
learn
not
only distribution
of
composite
quality, but also
responsibility
of a
constituent service
in
composite
quality
without actually observing the constituent’s
performance.
learns
quality
distribution of the
services
provides
how much each
constituent service
contributes to
overall compositionSlide11
Trust-Aware Service
Selection Models
Two Alternatives
Must be able to construct
model from
incomplete observations
Not
all
service qualities
are
observable
from the consumers’ point
of view
service quality attribute are represented as real numbers in interval
[0,
1]:
represent
observation
of a particular quality of service instance
d
at time
t
,…,
)
Slide12
Service Composition
Bayesian Model
P(T
)
P
robability
of obtaining
satisfactory
quality
from service T
Trust
Composite
S
ervice
atomicSlide13
Service Composition
Bayesian Model
C
onditional
probability table associated with each node provides a basis for
determining how
much responsibility to assign
to constituent services
Conditional
probabilities
represent level
of
trust consumer places
in
constituent services
in
composition
Slide14
Service Composition
Dealing with Incomplete Data
model
variables
may not be
observable
data is often incomplete
Variables w/o data considered
latent variable
Expectation Maximization (EM
)
is used to
optimally estimate
distribution parameters which
are then used to calculate the expected values of
latent variablesSlide15
Service Composition
Dealing with Incomplete Data
Example:
Travel
service
depends hotel
service
Consumer
observes that
has reliability 1 at time-step t
but does not
observe the reliability of
at time
So,
expected
reliability
of
,
can be used as nominal observation, i.e. ,
Completed
data
,
can
be used as the
observation in M step to update the parameter estimates using Bayesian inference. New parameter estimate can
be calculated by the posterior mean of
The
E
and M
steps are executed iteratively until the estimation
converges.
Slide16
Service Composition
Beta-Mixture Model
Superposition
of
multiple Beta probability
density
components, representing multiple subpopulations
E
ach mixing coefficient is
an indicator of
corresponding
component’s
responsibility
,
i.e., how
much contribution
component
makes toward
composite quality
Mixture dist. is governed by two parameters:
Slide17
Service Composition
Beta-Mixture Model
Mixture
distribution estimated by
maximizing log-likelihood
function using
EM
algorithm
: binary
latent variable,
indicating
whether an observation
is from component
k
. Exactly one of the
equals 1;
rest
are
zero.
Slide18
Service Composition
Beta-Mixture Model Estimation
EM Algorithm Steps
EM
is a
sequential
online
learning
algorithm
: it is repeated whenever
the consumer makes new
observations.Slide19
Experimental EvaluationsSlide20
Composition Operator
Service Quality Metrics and Interaction
Types
SWITCH
chooses exactly one of its children based on a predefined multinomial distribution
simulates composite quality based on one children
MAX
composes quality by inheriting from child with highest quality value
relates to latency for flow.Slide21
SWITCH
chooses exactly one children based on predefined multinomial distribution
simulates
composite quality based on one children
MAX
composes quality by inheriting from child
w/ highest
quality value
r
epresents
latency
for flow
MIN composes quality by inheriting from child with lowest qualitythroughput for sequenceSUM yields composite quality value as sum of quality values obtained from all childrenrelates to throughput for
flow
PRODUCT
yields
composite
quality value as
product
of
quality
values obtained from all
children.
relates
to
failure for flowSlide22
Experimental ResultsBayesian
AppraochSlide23
Composite Service C Trust Estimation
SWITCH OperatorSlide24
Composite Service C
Conditional
Trust
(SWITCH Operator)
Good Service
Bad ServiceSlide25
Bayesian vs. Naïve
Prediction
Errors
(80% missing data)Slide26
Conditional
Trust
in
Composite Service
MAX
MIN
40% data
m
issingSlide27
Dealing with Dynamic BehaviorSlide28
Random
Walk Service
Cheating
Constituent ServiceSlide29
Actual and Estimated Parameters Slide30
Estimated Beta-mixture
& Actual Distribution
and samples of trust
(
SWITCH composition)
Beta-mixture learns accurate distributions of both component services.
One provides good service (left peak); the other provides bad service (right peak).Slide31
Kolmogorov-Smirnov Test
FCM-MM vs.
B
eta-mixtureSlide32
Prediction Error
Nepal
et al.
vs. Beta-mixtureSlide33
P
owerful means of estimating quality distribution of a composite service w/o knowing quality
of
constituents
A
ccurately estimates responsibilities
of each constituent
service
Limitations
Difficult to learn component distributions when composite
distribution is
unimodal
. Accuracy may be improved if constituent services qualities are partially observable. Difficult to learn constituent services that rarely contribute due to lack of evidence; beta-mixture can correctly identify those services. Cannot track dynamic behavior.
Beta Mixture ModelSlide34
Limitations
lack
of unconditional trust in the constituent
services
a
ssumption
of
a least
partial observability
Bayesian ModelSlide35
Key features
Two probabilistic models for trust-aware service selection and composition
can handle variety of service composition patterns
Can
capture relationships between qualities of service
offered by
composite
service and
qualities
offered by its
constituents
Trust is
learned sequentially from directed observations then, combined with indirect evidence in terms of service qualitiesCan handle incomplete observationsSummary Slide36
Key features
Each consumer must monitor quality attributes of services it interacts with & maintain own model local knowledge
Model evaluation technique: simulation
Future r
esearch idea
Apply
Structural EM
, instead
of parameter
estimation, to learn
not
only trust information but also service dependency graph structure: learned structure can be used as a basis for suggesting new service compositionsSummary