BPM meets Semantic Web

Jörg Nitzsche | Daimler AG/IAAS, University of Stuttgart
Tammo van Lessen | IAAS, University of Stuttgart

BPM meets
Semantic Web

About Us
Tammo van Lessen Jörg Nitzsche

 Independent Consultant  Enterprise Architecture
 Researcher & PhD Student Management
IAAS, University of Stuttgart Daimler AG

 Committer & PMC Member  Researcher & PhD Student
Apache ODE IAAS, University of Stuttgart

Prof. Scheer on BPM 3.0
“Although semantic BPM currently
only exists in the research labs of
universities and a limited number
of BPM technology providers, such
as IDS Scheer, this technology is
key to the future of BPM. For it to
succeed, organizations must have a
high level of BPM maturity, but
given the rapid progress made
recently by BPM at the technology
and organizational levels, it can
only be a matter of a few years
before semantic BPM becomes
reality.” [1]
[1] Scheer, A.W., Klueckmann, J.: „The Future of BPM starts now!”, ARIS Expert Paper, 2009.
http://cdn.ariscommunity.com/media/pdf/BPM_-_Scheer_Klueckmann_-_Future_of_BPM_-_en.pdf

The Talk Today
• Overview
• Semantic Business Process Management
• Problem Statement
• sBPM Lifecycle
• Modelling
• Configuration
• Execution
• Analysis
• Conclusion

Semantic Business Process
Management
(sBPM)

Semantic Matchmaking

Semantic Web Services

Matching Model Representations & Semantics

Querying the Process Space

12

What is SUPER?
SUPER Semantics Utilised for Process management
within and between EnteRprises

EU Integrated Project (IP), FP6
Program: Information and Society Technologies
Begin: April 2006
Duration: 36 months
Funding: ~16,4 Million €
19 Partners, ~60 Researchers

Consortium:
SAP, CEFRIEL, Etel, Hanival, IBIS Prof. Thome, IBM, IDS Scheer, iSOCO,
LFU Innsbruck, MIP, NUI Galway, Nexcom, Ontotext/Sirma, Open
University, Telefonica, Telekomunikacja Polska, Poznan University, TU
Eindhoven, University of Stuttgart

Semantic BPM Lifecycle
• Modelling – add semantic
(ontological) annotations to
business processes (BPMN):
enables discovery of appropriate
process fragments (auto-
completion), enables composition

• Configuration – map from the
business model to an executable
process specification

• Execution – process execution
(BPEL) with discovery of SWS
during runtime

• Analysis – monitor, analyse &
improve processes

sBP Modelling
• Processes are modeled in a graphical notation known to
business people (e.g. BPMN)
• Processes are semantically annotatated
• The semantics of process tasks and events are specified
explicitly by using ontologies
• Different kinds of ontologies needed
• Process Ontology
• Organizational Ontology
• SWS Ontology (WSMO)
• Domain ontologies

• Features during process modeling:
• Semantic discovery of already existing processes and process
fragments in the SBP repository
Auto-Completion
• Refinement of conceptual models via composition
• Semantic Verification

17

Semantics for the WWW

Web Services Semantic Web
Dynamic
UDDI, WSDL, SOAP Services
OWL-S, WSMO, SAWSDL

Static
WWW Semantic Web
URI, HTML, HTTP RDF, RDF(S), OWL, WSML

WSMO – The Web Service Modelling
Ontology
Semantic Description of the
requirements a client has:
•Capability (funtional)
•Interface (usage)

Provide the formally Semantic description
specified terminology of Web Services:
of the information •Capability (functional)
used by all other •Interface (usage)
components

Connectors between
components with mediation
facilities for handling
heterogenities

WSMO – Web Services (& Goals)
- Advertising of Web Service - quality aspects
- Support for WS Discovery - Web Service Management

Capability Non-functional Properties

functional description DC + QoS + Version + financial

client-service realization of
interaction interface functionality by
for consuming WS Web Service WS
aggregating
- External Visible Implementation other Web Services
WS
Behavior (not of interest in Web - functional
Service Description)
- Communication WS
decomposition
Structure - WS composition
- ‘Grounding’
Choreography --- Service Interfaces --- Orchestration

Attaching WSMO Goals to Processes
client-service - quality aspects
- Advertising of Web Service interaction interface - Web Service Management
- Support for WS Discovery for consuming WS
- External Visible Non-functional Properties
Capability Behavior
- Communication
DC + QoS + Version + financial
functional description Structure
- ‘Grounding’

</interactionActivity>
...
messageRef=“out-only:Out”/>
<input messageLabel=“out-optional-in:Out”
mep=“out-only”
<interactionActivity ...

Features during Process Modelling
• Semantic discovery of already existing processes
and process fragments in the SBP repository
Auto-Completion

• Semantic Verification

• Refinement of conceptual models via composition
• Composition queries the service repository for
appropriate services for each task of a process

• In case there is no service for a given task,
multiple services are combined

• Based on ontologically defined inputs and
outputs and ontological reasoning

Configuration options
• During configuration there are a couple of options to chose
from

• Discover WS during configuration
 Conventional BPEL is produced
• Endpoint references are fixed
• Endpoint references are discovered during runtime based on WSDL
interfaces
• Pros: mature technology
• Cons: limited flexibility

• Discover WS only during runtime based on semantic goal
descriptions
 The result is BPEL4SWS
• Pro: enhanced flexibility
• Cons: semantic middleware is only implemented prototypically

BPEL Process
BPEL Process
… Online Shop
WSDL Service
salesPLT
A

… orderItem

… B payBill

…

C getShippingDetails

…

BPEL for Semantic Web Services
BPELlight
Extension of BPEL 2.0
Removes dependencies on WSDL.
Describes Message Exchanges.
Independent of any IDL.
Grouping Activities to
Conversations and to logical partners
“Binding” to Services is done
during deployment.

BPEL4SWS
Defines such a “Binding” for Semantic Web Services.
Introduces an activity to perform data mediation
Proposes an API for SWS middleware to support long running service
interactions.
Uses SAWSDL for Lifting and Lowering.

BPEL4SWS Process
BPELlight Process
Aktivity definition
… with WSMO Online Shop
WSDL Service
WSMO Goal WSMO Service
A
Capabilities Capabilities orderItem
Choreography Choreography
…
Achieved
by
B
payBill

…
getShippingDetails
Aktivity C


C getShippingDetails

…

SBP Execution – Architecture
Tools SEE SBPELEE
Modelling Service Navigator
Tool Discovery
Build Time DB
Monitoring Service
Tool Selection

Mining Tool Service Runtime DB
Invocation

sRBE Tool
WS Gateway SWS Gateway Event Publisher Deployment Mgr

Semantic Service Bus (SSB)

SPAB/SWAB Binding Data LILO
Deployer Components Mediators
SWS Execution
Repository History

40

SBP Execution – Architecture
Tools SEE SBPELEE
Modelling Navigator
Tool
Build Time DB
Monitoring 1. Deployment
Tool
Black box 2. Execution of sBP
Mining Runtime DB
Tool 3. Monitoring
sRBE Tool


SWS Execution
Repository History

41

SBP Deployment
BPEL4SWSs

(SA-)WSDLs

Deployment Descriptor
Semantic Process
Artefacts Bundle WSMO Goals
(SPAB)
Partner WSMO WSs

Partner WSDLs

Lifting & Lowering Definitions

SBP Deployment (2)
Tools SEE SBPELEE
Tool Discovery
Build Time DB
Monitoring Service
Tool Selection

Mining Tool Service Runtime DB
Invocation

sRBE Tool


SWS Execution
Repository History

SBP Execution (vanilla WS)
Tools SEE SBPELEE
Tool Discovery
Build Time DB
Monitoring Service
Tool Selection

Mining Service Runtime DB
Tool Invocation

sRBE Tool


SWS Execution
Repository History

44

SBP Execution (SWS)
Tools SEE SBPELEE
Tool Discovery
Build Time DB
Monitoring Service
Tool Selection

Tool Invocation

sRBE Tool


SWS Execution
Repository History

45

SBP Execution – Generic LILO
WSMLreq WSMLresp

SBPELEE
SEE
XMLreq

WSMLreq Lifting & WSMLreqS
Lowering XMLreqS
WSMLresp Component
XMLrespS
XMLresp
WSMLrespS

XMLreqS
XMLrespS
-Supports orchestration of legacy WS (backward compatibility)
-Processes are provided as WS and SWS
-Virtual Dualism of BPEL4SWS variables WSs…
-SEE serves as ‘semantic adapter’ for legacy WS

46

SBP Execution – Monitoring
Tools SEE SBPELEE
Tool Discovery
Build Time DB
Monitoring Service
Tool Selection

Tool Invocation

sRBE Tool


SWS Execution
Repository History

47

Nexcom Prototype
B2B and Automated Decision Making
Implemented business process

49 Ivan Pavlov(Nexcom)

Nexcom Prototype
■ The customer enters data (name, address) requested
services and QoS parameters
■ The system:
■ Checks for available services ( B2B partners)
■ Extracts data about the customer from IT Systems
■ Reasoning is used to rate the services:
■ Customer credit rating and company policy are considered
■ Highest rated are the services which Nexcom is most willing to
offer for this customer
■ The customer receives offers, ordered by their rating
■ In the full implementation, the customer will see only the
“best” offers ????
■ The sales person manages the process:
■ Edits the ontologies
■ Monitors the process

Scenario
Platform Services Execution Environment
Lifting & SBPELEE SEE SWS Nexcom
Lowering Repository Ontology
Repository

SSB

Binding Components Nexcom Reasoning Services
HTTP Service1 Service2 ServiceN
…

…

Nexcom Legacy Systems Marketplace
Service Layer Services

Nexcom
Nexcom
Administrator
Clients GUI
GUI

SBP Analysis & Monitoring
• Semantic annotations can significantly improve the
results of process mining tools.
• Open World Assumption makes domain and organisational
models accessible.
• Ontologies and Mediators help to introduce terminology
transparency
• During runtime, semantic
model references can be
seen as “traces” that
allow for inferring
information about which
element of a business
model relates to a
technical, executable
model.

sBP Monitoring & Analysis Tools

Conclusions
•Semantics can help narrowing the gap
between business and IT…
•Modelling: composition, auto-completion,
validation
•Configuration/Execution: Improved Flexibility,
loose coupling
•Analysis: semantic queries based on ontologies
•…but: requires additional effort
•Prototypes for all phases of the lifecycle exist

Thank you for your attention!

Questions?

Further information: http://ip-super.org

BPM meets Semantic Web

More Related Content

What's hot (19)

Similar to BPM meets Semantic Web (20)

More from Tammo van Lessen (11)

Recently uploaded (20)

BPM meets Semantic Web