Filtering & Selecting Semantic Web Services with Interactive Composition Techniques
By
Evren Sirin, Bijan Parsia, and James Hendler
Presenting By : Mirza Tania Nasreen Mohammad Hasan
2
Outline of the Presentation
Main Idea of the Paper Works done by Authors Relevant Terms Comparisonal Work Flow Technology Used Implementation details Conclusion
4
Why Composition ???
Goal: applications will be able to communicate each other very flexibly to achieve some combined and new functionality from existing ones.
Present Condition: failed to achieve this goal
the expected ability to compose web services has not been achieved yet
5
Works Done by the Authors
Developed goal-oriented interactive composition approach
Implemented In a Prototype system
Help the users to dynamically compose web services from internet with filtering capabilities
6
Web Service
Web Service is a software system designed to support interoperable Machine to Machine interaction over a network. [W3C]
A web service is a collection of protocols and standards used for exchanging data between applications or systems. [Wikipedia]
No Semantics in WSDL
Current State of SOA
8
Semantic Web Service
World Wide Web
Semantic Web
a services whose properties, capabilities, interfaces, and effects are encoded in an unambiguous, machine-understandable form.
Web Service
Semantic Web Service
Web Service + Semantics = Semantic Web Service
[ HOW & WEHERE ]
[ WHAT & WHY ]
9
Why Semantics ????
(Discovery) Neither WSDL nor UDDI allows for software to determine what a Web service offers to the client. A Semantic Web service describes its properties and capabilities so that software can automatically determine its purpose.
(Invocation) A Semantic Web service provides a descriptive list of what an agent needs to be able to do to execute and fulfill the service. This includes what the inputs and outputs of the service are.
(Monitoring) services can interoperate with each other seamlessly and can combine results for a valid solution.
Dynamic Composition
10
Why Semantics? An Example
Semantic Annotation is the main concept behind Dynamic Composition
DictionaryTranslator
word translated word word meaning
French to English Dictionary
French word translated English word meaning
How & Where
WSDL
How & Where
WSDL
string string string string
WS 1 WS 2
+What & Why
+What & Why
11
How Semantics ????
UDDI
Dynamic Composition
WSDL
SOAP
Standards for Service Directory
Standards for Description
Standards for Messaging Protocols
First System to directly combine the OWL-S semantic service descriptions with actual invocations of the WSDL descriptions.
WSDL + OWL-S = Semantic Web Service Description
12
Major Focuses of the Paper
Filtering and Selecting Semantic Web Services with
Interactive Composition Techniques
Filtering Selecting Interactive Composition
Techniques
13
Filtering
When a service goes into the composition, this service’s information about input, output, preconditions, and effects (IOPE) serves to automatically filter the services whose outputs are incompatible with the current selection.
Filter is designed to examine each input or output request for certain qualifying criteria and then process or forward it accordingly.
Filtering helps to determine the service that best fits user’s/clients personal preferences.
Filtering for their tool done by using matchmaking algorithms
15
Interactive Composition
A successful, executable composition correctly combines a set of compatible components to achieve the composition’s overall goal.
At each step, their system adds a new service to the composition and filters further possibilities based on the current context and user decisions.
Gradually generates the composition with a forward or backward chaining of services.
Filtering + Selecting = Interacting Compositing Techniques
16
Step-by-step Composition
Fixing a GOAL
Select 1st web Service
Filtering
Filtering
Select 2nd web Service
Compose 2 web services
RESULT
Auto. Filtering
List of Services
Selecting
Auto. Filtering
List of Services
Selecting
Monitoring
Selection
Manual Composition Dynamic CompositionPartial Automation of Composition
17
Creating Semantic Service Description – OWL-S OWL : Enables greater access only to content
OWL-S ( formerly DAML-S): Enables greater access to the Web Services
OWL-S partitions a Web Service’s description into three components :
1. Service Profile
- IOPE parameters
- service parameter
2. Process Model
-Atomic Process
-Composite Process
3. Grounding
-Mapping from OWL-S to WSDL
18
Translation from WSDL to OWL-S WSDL Operation ≡ OWL-S Atomic Process WSDL message parts ≡ OWL-S Process’s Parameters
Difficulties with Type Conversion :
Message parts are described by XML Schema data types OWL itself permits only a subset (constrained range) of
XML Schema data types (integers / strings). OWL references data types by URI No canonical way in XML Schema to determine a URI for
a complex data type
Preferred Solution : Parameter types of OWL-S services be OWL classes
19
Translation from WSDL to OWL-S cont’d…
Author’s Solution: Treated the WSDL-supplied types as descriptions of the service
parameters i.e. the serialization of the values the process actually uses. Extended the OWL-S grounding to include marshaling and
unmarshaling functions using XSLT Unmarshalling function : XML Schema type to an RDF graph
serialized in the RDF/XML exchange syntax That graph encodes the relevant assertions about the individual,
which becomes the actual input to the service
Difficulties: It’s difficult to write XSLT that can handle all the legal
serializations of a given RDF graph.
21
Implementation 4 Types of IOPE Matching:
Exact: If advertisement A and request R are equivalent concepts
PlugIn: If request R is a subconcept of advertisement A
Subsume: If request R is a superconcept of advertisement A
Fail: No match.
22
Matching IOPE
Only IO was used for matching Specifications of PE are still an open OWL-S issue Exact and Plug-In matches between the parameters of
ServiceProfiles yields useful results
Returns an ordered list
23
Matching Service Parameters
To get rid of long list of available choice Service names themselves might not contain
enough information Means of introducing more user constraint Applies the result of this new query to the
previous result set
25
Generating and Executing Composed Services
Generation : Each Composition ≡ OWL-S CompositeProcess It can also be advertised, discovered, and composed
with other services Generates exactly such a CompositeProcess
description Creates the corresponding ServiceProfile
Execution : Invoking each individual services and passing the
data Client program serves as the central control authority
26
Improvement and further automation
Strong need for a suitable set of service descriptions of sufficient and compatible detail to support
Converting the IO type descriptions from XML Schema data types to OWL classes
Removing human interaction from the loop by integrating a planner
Introduce machine learning – better and preferred suggestions
27
Conclusion
Semantic Web
Semantic Web Service
Maturity Hill
Web Service
Full Automation
Partial Automation of Composition
28
References
Filtering and Selecting Semantic Web Services with Interactive Composition Techniques. IEEE Intelligent Sytems. 19(4): 42-49 (2004).
Semantic Web Services. IEEE Intelligent Systems, Volume: 16 Issue: 2, March-April 2001, Pages: 46 -53
Semantic Web Service Architecture — EvolvingWeb Service Standards toward the Semantic Web. American Association for Artificial Intelligence (www.aaai.org).
Bringing Together Semantic Web and Web Services. Proceedings 1st International Semantic Web Conference (ISWC 02), 2002.
http://www.w3.org/Submission/OWL-S/