Integrating an Enterprise Architecture Ontology in a Case-based Reasoning Approach for Project Knowledge

CTI # 14575.1 PFES-ES
Andreas Martin - FHNW
Integrating an Enterprise Architecture
Ontology in a Case-based Reasoning
Approach for Project Knowledge
Andreas Martin, Sandro Emmenegger and Gwendolin Wilke
Integrating an Enterprise Architecture Ontology in a Case-based Reasoning
1

Table of contents
1. Introduction
1. CTI founded Research Projects
2. [sic!]
3. The application partner
2. Project Goal and Application Scenario
3. The Approach
4. Implementation
5. Conclusion & Future Work
2

Introduction
CTI founded Research Projects / [sic!] / the application partner
3

CTI founded Research Projects
 The Commission for Technology and Innovation (CTI)
promotes projects in applied research and development
between centres of higher education and companies.
 The Swiss Confederation founds 50 % of the total costs.
 The application partner must cover at least 50 % of the total
costs - the cash contribution must equal at least 10% of the
federal contribution.
 This work was supported in part by the CTI under Grant
14575.1 PFES-ES.
4

 This work is an outcome of the research project [sic!].
 [sic!] stands for software integration using ontology-based
case-based reasoning.
 Start: September 2012
 End: January 2015
 CTI – founding (50%): CHF 285’000.-
 Application partner: ELO Digital Office CH AG
5

The application partner – ELO Digital Office CH AG
 The ELO Digital Office CH AG is a subsidiary of the ELO Digital
Office GmbH, which has its headquarters in Stuttgart
(Germany).
 ELO develops and sells software solutions in the areas of
electronic document management, digital archiving and
workflow management - Enterprise Content Management
(ECM).
 ELO Digital Office CH AG is an own legal entity and acts on
the Swiss market.
 ELO Digital Office CH AG has an extensive network of local
partners.
6

Project Goal and Application Scenario
7

Problem
 ELO is an international company with branch offices and
distribution partners all over Europe…
 …project knowledge is distributed over different people…
 …and also over different teams in different locations.
 For a project worker, it is of vital importance to have access
to other people’s historic project knowledge and
experience.
8

Problem and Goal
 Current Situation: ELO uses enterprise content management
and document management systems, workflow
management systems, a user forum, as well as a centralized
project database.
 Problem: An all-embracing management of historic project
knowledge based on problem descriptions is not available.
 Goal: Improving and optimizing experience management
process …
 … by implementing an ontology-based case-based reasoning
system.
9

Application Scenario
 The application scenarios are elicited from our business
partner ELO Digital Office CH AG.
 An IT- project usually consists of the following three phases:
 In every phase certain project knowledge is needed.
 From other people… / about certain technical issues… / etc.
 We focus on the sales phase and we derived two exemplary
application scenarios:
 Application Scenario 1: Answering a customer’s questionnaire.
 Application Scenario 2: Searching for a module expert.
10
Sales Implementation
Operation and
Maintenance

Application Scenario - Overview
11
Operation and
Maintenance

12
Operation and
Maintenance

13
Operation and
Maintenance

Application Scenario 1
Answering a customer’s questionnaire.
 In the tendering part of the sales
phase of an ECM project, a
detailed offer is assembled.
 The offer is based on the
customer’s specifications and
requirements catalogue, which is
usually handed out as a
questionnaire.
14

Typical question:
 Is the integration of an ELO
product or module possible with a
customer’s system component?
If the person in charge does not know
the answer…
 …it may be helpful to retrieve
similar project experience and
related documentation.
15

The information need of a person trying to
answer a question includes details of the
integration such as the question
 if integration is possible,
 if customization is necessary (and
possible),
 if additional programming effort is
necessary,
 if there are function parameters, or
 if there are functionality constraints.
 Example: “Does the ECM/DSM software
support archiving the MS Exchange 2007
journal?”
16

Searching for a module expert.
 Software integration is often a
complex process, where usually
different employees are in charge
of different subtasks.
 In particular, employees usually
specialize on different aspects of
the integration process or on
different ELO modules. E.g., a
person with a strong SAP
background might be an “ELO SAP
module” specialist.
17

Searching for a module expert.
 it may be helpful to retrieve
similar historic projects in order to
find an module expert.
 In this scenario, the information
need of the project manager is to
find a module expert who has
experience with integrating the
customer’s system component in
question.
18

The Approach…
…and Related Work
19

Three basic underlying research topics…
1. Case-based Reasoning
(CBR)
2. Enterprise Architectures
3. Enterprise Ontologies
… why using them?
 This work is an outcome of
a «Design Science
Research» instantiation:
20
Design science research cycles adapted from (Hevner and Chatterjee, 2010; Hevner et al., 2004)

The Approach
1. Case-based Reasoning (CBR)
21

What is Case-based Reasoning (CBR)
 CBR can be seen as “reasoning by remembering”…
 and it is a technically independent methodology to humans
and information systems.
 “Case-based reasoning is both […] the ways people use cases
to solve problems and the ways we can make machines use
them”.
 Two central elements:
1. the CASE
2. the CBR- CYCLE (& CASE- BASE)
(a) SIMILARITY / (b) Adaptation / (c) Evaluation / (d) Learning
22

CASE
 Traditional CBR terminology: a case consists of a
problem space (problem items /
descriptions) that is used for describing a
certain solution space (solution items).
 Bergmann’s CBR terminology: a case consists of a
case characterization space that is used for
describing a certain lessons space (derived
from “lesson learned”).
 Our CBR terminology: a case consists of a case
characterization (sometimes called
metadata) that is used for describing a
certain case content (sometimes called
lesson).
23
Characterization
Content
CASE
That is a concession to the
business needs (relevance) –
familiar and domain-oriented
wording.

CBR- CYCLE (& CASE- BASE)
 Retrieve the most similar cases from the
knowledge base (case-base containing
previous cases) based on the problem
description of the new case (problem case)
using a similarity mechanism.
 Reuse the knowledge in the retrieved case(s)
in order to solve the current problem – adapt
the historical knowledge to the new problem
(adaptation).
 Revise and test the suggested solution e.g. by
evaluating it under the real world problem
(evaluation).
 Retain useful experience (past solutions and
failures) for future reuse and store a new
case in the knowledge base (case learning).
24
Based on: A. Aamodt and E. Plaza, “Case-Based Reasoning : Foundational Issues , Methodological Variations , and System Approaches,”
Artificial Intelligence Communications, vol. 7, no. 1, pp. 39–59, 1994.

Why Case-based Reasoning (CBR)?
Findings
 Our domain expert focus group (ELO people)
 are experts in Enterprise Content Management (ECM).
 They are using the latest technology on the market (their own
software) for project management.
 They have strong expertise in IT and business consulting.
 Our main finding is: They are thinking in CASES and
METADATA.
25

Why Case-based Reasoning (CBR)?
Relevance
1. Requirements from business:
 A way to gather information and lessons from project- work
including metadata and data. → CASE (vocabulary)
 A knowledge/data base -> Case- BASE
 A way to retrieve similar cases -> SIMILARITY
 A generic management process or method -> CBR- CYCLE
 A prototypical implementation as IT system -> Case-based Reasoning
Application
26

Why a new Case-based Reasoning (CBR) Approach?
Relevance
2. Requirements from business:
 Focus on standardized methodology and technology.
 ELO wishes to easily extend existing models and reuse
existing knowledge about the enterprise.
27

The Approach
2. Enterprise Architectures
28

How to re-use existing knowledge?
What can be used and what is available?
29

Enterprise Architecture
Knowledge about the Enterprise
 Enterprise Architectures (AE) are a way to model the relevant
aspects of an enterprise and interdependencies between business
and information systems.
 An Enterprise Architecture (AE) is
 “[…] a coherent whole of principles, methods and models that are used
in the design and realisation of an enterprise’s organisational structure,
business processes, information systems, and infrastructure” (Lankhorst
2009, p. 3).
 Example Enterprise Architecture Frameworks (EAF):
 Zachman Framework
 The Open Group Architecture Framework (TOGAF)
 ArchiMate
30
Lankhorst, M., 2009. Enterprise Architecture at Work. Berlin, Heidelberg: Springer Berlin Heidelberg.

ArchiMate®
31
The Open Group, “ArchiMate® 1.0 Specification,” 2009. [Online]. Available: http://pubs.opengroup.org/architecture/archimate-
doc/ts_archimate/.
ArchiMate is a technical standard from The Open Group
and is based on the concepts of the IEEE 1471 standard.

ArchiMate® - Example
 Business layer
 Application layer
 Technology layer
32
The Open Group, “ArchiMate® 1.0 Specification,” 2009. [Online]. Available: http://pubs.opengroup.org/architecture/archimate-
doc/ts_archimate/.

The Approach
3. Enterprise Ontologies
33

How re-use existing knowledge?
34

Enterprise Ontologies and Semantic Technologies
 “An ontology is a formal, explicit specification of a shared
conceptualisation” (Studer, 1998, p. 184)
 “The main purpose of an enterprise ontology is to promote
the common understanding between people across
enterprises, as well as to serve as a communication medium
between people and applications, and between different
applications” (Leppänen, 2007, p. 273)
 Semantic Technologies:
35
http://smiy.wordpress.com/2011/01/10/the-common-layered-semantic-web-technology-stack/

 ArchiMEO is an enterprise ontology
based on ArchiMate and is extended
with selected concepts from other
enterprise ontologies.
 ArchiMEO is implemented using
RDF(s) and OWL.
 ArchiMEO has been developed by
several team members of the FHNW
Information and Knowledge
Management Research Group (IKM).
 ArchiMEO is licensed under a
Creative Commons Attribution-
ShareAlike 3.0 Unported License.
 ArchiMEO is available for download
as TTL- files (Terse RDF Triple
Language) under:
ikm-group.ch/archimeo
36
ArchiMate is a technical standard from The Open Group.
RDF(S) / OWL is a W3C standard.
The enterprise ontology ArchiMEO is based on ArchiMate.

The Approach
4. Integrating an Enterprise Architecture Ontology in a Case-based
Reasoning Approach for Project Knowledge
37

The Approach
38

Why a new Case-based Reasoning (CBR) approach?
Rigor
 Our state of the art analysis has shown that there is a
potential for new ontology-based CBR approach, which uses
an enterprise architecture formalized in an enterprise
ontology.
39

The Approach applied to Application Scenario
40

Implementation
41

Implementation
 [sic!] is an iterative research project.
 The results presented here are the outcome of the first
implementation iteration (Prototype I).
 Prototype I: Case Retrieval
 Case-based Reasoning Ontology
 CBR- Retrieval Component
 User Interface
 Similarity- Functions
 Prototype II: whole CBR-Cycle
42

Implementation - Ontologies
 ArchiMEO: the Enterprise Ontology
as basis.
 Case Ontology: selected elements
from the Case Management Model
and Notation (CMMN).
 Similarity Ontology: retrieval
mechanism.
 CBR Ontology: extends the Case
Ontology and the Similarity
Ontology for CBR specific needs.
 Project Ontology : contains concepts
that are specific for project related
use cases.
 ELO Domain Ontology: domain
specific ontology that contains
knowledge of ELO.
43

Implementation - Similarity
44

Implementation - Similarity
45
What will be compared?
Case Characterization
Content
«New case»
Content
«case 1»
Content
«case 2»
Content
«case X»
Compare

Implementation
Case characterization stored in ontology
46
System: Third-party systems that will be connected (e.g. ERP System)
Requirement: Requirement and Solution (e.g. Archiving)
Module: ELO – module (e.g. Backup)
BusinessActor: Module expert

Implementation
Similarity - Weighting
47
Application Scenario 1: Answering a customer’s
questionnaire.
• Which third-party system should be integrated?
• Which requirement should be fulfilled?
• Which ELO- module would be worth considering?
Application Scenario 2: Searching for a module
expert.
• Which ELO system is relevant?
• Who is an expert for a specific ELO module?

Implementation - Similarity - Annotations
 Similarity functions:
(a) levenshtein: minimal number of edit operations when transforming one string to another.
(b) version: custom function for comparing versions.
(c) average: average of numbers. (d) equals: is one string identical to another.
48
cbr:Case
elo:System
elo:Module eo:Person
hasSystem
hasModule hasExpert
name
version
name
role
levelrdfs:label
ObjectPropertySim
weight: 1
simFunction: average
ObjectPropertySim
weight: 5
AnnotationPropertySim
weight: 3
simFunction: equals
annotationProperty: label
language: en
ObjectPropertySim
weight: 2
DatatypePropertySim
weight: 2
simFunction: levenshtein
DatatypePropertySim
weight: 1
simFunction: version
DatatypePropertySim
weight: 1
simFunction: equals
DatatypePropertySim
weight: 3

Implementation - Similarity - Query Case
49
cbr:Case
elo:System
hasSystem
hasModule hasExpert
name
version
name
role
levelrdfs:label
ObjectPropertySim
weight: 1
ObjectPropertySim
weight: 5
weight: 3
simFunction: equals
language: en
ObjectPropertySim
weight: 2
DatatypePropertySim
weight: 2
DatatypePropertySim
weight: 1
DatatypePropertySim
weight: 1
simFunction: equals
DatatypePropertySim
weight: 3
cbr:Case
elo:System
hasSystem
hasModule hasExpert
name
levelrdfs:label
«_queryCase»
«_querySystem»
«mySQL»
«_queryModule»
«Backup»
«_queryExpert»
«Expert»
Query Case

cbr:Case
elo:System
hasSystem
hasModule hasExpert
name
version
name
role
levelrdfs:label
ObjectPropertySim
weight: 1
ObjectPropertySim
weight: 5
weight: 3
simFunction: equals
language: en
ObjectPropertySim
weight: 2
DatatypePropertySim
weight: 2
DatatypePropertySim
weight: 1
DatatypePropertySim
weight: 1
simFunction: equals
DatatypePropertySim
weight: 3
cbr:Case
elo:System
hasSystem
hasModule hasExpert
name
role
levelrdfs:label
«Case2»
«case2System»
«MySQL»
«case2Module»
«Backup»
«case2Expert»
«Programmer»
«Beginner»
version
«5.1»
cbr:Case
elo:System
hasSystem
hasModule hasExpert
name
role
levelrdfs:label
«Case1»
«case1System»
«Oracle»
«case1Module»
«Barcode»
«case1Expert»
«TechConsultant»
«Expert»
version
«11g»
Implementation - Similarity - Cases
50
Case 1 (in Case Base) Case 2 (in Case Base)

Class Instance Property Weight Function Sim #1 Sim #2 Sim #3
Case
“_queryCase”
hasSystem = ”_querySys”
hasModule = ”_queryMod”
“case1”
hasSystem = ”case1Sys” 1 average 0.17
0.46
hasModule = ”case1Mod” 5 average 0.52
“case2”
0.66
System
“_querySys”
name = “MySQL”
version = “”
“case1Sys
name = “Oracle” 2 levenshtein 0.17
0.17
version “11g” 1 levenshtein
“case2Sys
name = “MySQL” 2 levenshtein 1.0
1.0
version = “5.1” 1 levenshtein
Module
“_queryMod”
label = “Backup”
hasExpert “_queryExp”
“case1Mod”
label = “Barcode” 3 equals 0.2
0.52
hasExpert = “case1Exp” 2 average 1.0
“case2Mod”
label = “Backup” 3 equals 1.0
0.6
hasExpert = “case2MExp” 2 average 0.0
Employee
“_queryExp”
role = “”
level = “Expert”
“case1Exp”
role = “TechConsultant” 3 levenshtein
1.0
level = “Expert” 1 equals 1.0
“case2Exp”
role = “Programmer” 3 levenshtein
0.0
level = “Beginner” 1 equals 0.0
Case
“_queryCase”
“case1”
hasSystem = ”case1Sys” 1 average
hasModule = ”case1Mod” 5 average
“case2”
System
“_querySys”
name = “MySQL”
version = “”
“case1Sys
0.17
“case2Sys
1.0
Module
“_queryMod”
“case1Mod”
label = “Barcode” 3 equals
hasExpert = “case1Exp” 2 average
“case2Mod”
label = “Backup” 3 equals
hasExpert = “case2MExp” 2 average
Employee
“_queryExp”
role = “”
“case1Exp”
1.0
“case2Exp”
0.0
Case
“_queryCase”
“case1”
“case2”
System
“_querySys”
name = “MySQL”
version = “”
“case1Sys
0.17
“case2Sys
1.0
Module
“_queryMod”
“case1Mod”
0.52
“case2Mod”
0.6
Employee
“_queryExp”
role = “”
“case1Exp”
1.0
“case2Exp”
0.0
Case
“_queryCase”
“case1”
0.46
“case2”
0.66
System
“_querySys”
name = “MySQL”
version = “”
“case1Sys
0.17
“case2Sys
1.0
Module
“_queryMod”
“case1Mod”
0.52
“case2Mod”
0.6
Employee
“_queryExp”
role = “”
“case1Exp”
1.0
“case2Exp”
0.0
Implementation
Similarity - Computation
51
cbr:Case
elo:System
hasSystem
hasModule hasExpert
name
version
name
role
levelrdfs:label
ObjectPropertySim
weight: 1
ObjectPropertySim
weight: 5
weight: 3
simFunction: equals
language: en
ObjectPropertySim
weight: 2
DatatypePropertySim
weight: 2
DatatypePropertySim
weight: 1
DatatypePropertySim
weight: 1
simFunction: equals
DatatypePropertySim
weight: 3
elo:System
eo:Person
name
version
name
role
level
elo:Module
rdfs:label
cbr:Case
elo:System
elo:Module
hasSystem
hasModule
46%
66%

System architecture
 Technology:
 Apache Jena: open source
Java framework for building
Semantic Web applications
 OpenDolphin: open-source
library for a lightweight
remote model-view-
controller separation.
 JavaFX: GUI framework
 TopBraid Composer: an
ontology engineering
software (paid)
52

Conclusion & Future Work
53

Conclusion I
Research perspective:
 It is possible to retrieve historic project knowledge from a
knowledge base using case based reasoning.
 The novelty is the in-ontology approach that embeds the
knowledge of the enterprise architecture ArchiMate in the
CBR system using the W3C conform formalization ArchiMEO.
 The similarity mechanism uses relations in the ontology in
order to calculate the overall similarity of a query case to a
historical case.
54

Conclusion II
Business perspective
 Standardized methodologies and technologies:
 Enterprise concepts and relations are based on ArchiMate.
 «ArchiMate is a technical standard from The Open Group and is based on the
concepts of the IEEE 1471 standard»
 ArchiMate is documented and not proprietary.
 ArchiMEO is a RDF(S)/OWL ontology.
 «RDF(S) / OWL is a W3C Semantic Web standard»
 No proprietary database technology.
 CBR- Similarity uses the SPARQL Inferencing Notation (SPIN) on
persistence layer
 «SPIN is a SPARQL-based rule and constraint language for the W3C Semantic Web.
SPIN is W3C Member Submission and open specification»
 No proprietary code on persistence layer and inference support on business logic
layer.
55

Future work:
 Implementation perspective:
 Current - Prototype 1: Case Retrieval
 Automatic Retrieve (Similarity)
 Prototype 2: CBR- Cycle
 Automatic Retrieve (Similarity),
 Manual Reuse (Adaptation), Revision (Evaluation),
 Automatic Retain (Case Learning - adding to case-base)
 Prototype 3: Case Adaptation & Learning
 Automatic Retrieve (Similarity),
 Semi-automatically Reuse (Adaptation - OWL/Rule Reasoning & ML) ,
 Manual Revision (Evaluation),
 Automatic Retain (Case Learning & Ontology Learning - adding to elements to domain ontology,
OWL/Rule Reasoning)
 Research perspective:
 Enhance usability and retrieval in Ontology-based CBR using NLP- technology
 Adaptation algorithms for the Ontology-based CBR [sic!] approach
 Further sophisticated similarity algorithms for the Ontology-based CBR [sic!]
approach
56

57
 Knowledge Work and Case-based
Reasoning
 Agile Business Process Management
and Workflow Systems
 Enterprise Software Engineering,
Architectures & Development
 Semantic Technologies
 Enterprise Architectures / Ontologies
 Information & Knowledge Management
 Computational Linguistics and Natural
Language Processing
You can find me on:
 linkedin.com/in/andreasmartinch
 andreasmartin.ch
Team member of the FHNW
«Information and Knowledge
Management Research Group» IKM.
Project lead of the CTI- project
«Software Integration using Ontology-
based Case-Based Reasoning» [sic!].
Contributor to the ArchiMate based
enterprise ontology ArchiMEO.
Always interested in a research collaboration in the following fields:

Integrating an Enterprise Architecture Ontology in a Case-based Reasoning Approach for Project Knowledge

More Related Content

Similar to Integrating an Enterprise Architecture Ontology in a Case-based Reasoning Approach for Project Knowledge (20)

More from andreasmartin (8)

Recently uploaded (20)

Integrating an Enterprise Architecture Ontology in a Case-based Reasoning Approach for Project Knowledge