semantic business rules for data validation
TRANSCRIPT
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
Simon Schlosser
Collaborative business partner data management Focus on semantic data validation
Manchester, June 25, 2015
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 2
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
Motivation
Agenda
Corporate Data League
Approach and Application: Semantic Business Rules
Discussion
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 3
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
Companies from several industries are facing the same
challenges in business partner management
DATA MANAGEMENT SHARECONOMY
Ever-growing regulation and
Compliance rules
Operational excellence – or
“do more with less”
Rise of professionally
digitized frauds
Find ways to improve process
performance with less head count
Digitize manual processes to reduce
head count and risk
Learn from others to avoid “reinventing
the wheel”
Advancement in hacker capabilities
Scammers perform multi-channel attacks
(e.g. hacked email and phone)
Professional frauds are run again multiple
companies
Provide upstream transparency and certificates
to fulfill regulation and consumer demands
Provide downstream transparency to comply to
e.g. blacklists and embargos
Provide high-quality data to fulfill legal reporting
requirements
Share updated business
partner data
Share fraud warnings
with forged data
Share certificates
and blacklists
Share updated business
partner data
Share fraud warnings
with forged data
Share certificates
and blacklists
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 4
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
Collaboration and cloud technology are the major pillars of
the Corporate Data League (CDL)
Corporate Data League*
The CDQ Team
Manages the CDL community as a
neutral moderator
Operates CDL cloud and services,
Monitors data quality and
process performance
Provides up-to-date reference data
(e.g. blacklists, business rules)
The CDL Cloud
Protects all CDL data by
state-of-the-art IT security
concepts
Keeps all data in a protected
cloud hosted at Swisscom in
Switzerland
The CDL Members
Share updates of business partner data
Double-check updates to assure data quality
Share compliance information such as uncovered
banking data frauds
Review and revise CDL metadata such as business
rules and blacklists
Share best practices for processes and organization
* CDL members and interested companies
CDL Cloud read
write
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 5
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
Motivation
Agenda
Corporate Data League
Approach and Application: Semantic Business Rules
Discussion
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 6
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
CDL cloud services provide a number of benefits across the
entire business partner data lifecycle
Lookup
Required information:
Name, country, locality
Add details
e.g. legal form and address,
tax number, and banking data
Check compliance
Check data against
CDL compliance database
Get regular updates
Call CDL cloud services
to receive updates
Higher data quality Less maintenance Less compliance risks
Lookup business
partner data in CDL
database
Cleanse addresses
Translate addresses
Validate data
Lookup banking data
Match against
blacklists
Check for banking
fraud warnings
Pull updated
business partner IDs
Pull updates per
record
Skip maintaining details and just copy complete
and up-to-date data including compliance checks Business partner
found in CDL database
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 7
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
The Corporate Data League is an unique community for
collaborating in business partner data management
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 8
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
Shared perceptions about the domain and its rules are a
prerequisite for collaboration
How to ensure a common
understanding of the data to
be shared?
How to ensure the quality of
the data?
How to collaboratively define
a shared business rule set for
business partner data?
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 9
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
Motivation
Agenda
Corporate Data League
Approach and Application: Semantic Business Rules
Discussion
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 10
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
The approach integrates the data model, business rules
and actual data instances into one common model
Represent the
domain of the
Corporate Data
League
Concepts
1
• Constraints:
It is necessary that a business partner has
exactly one name
It is necessary that a German post code
consists of exactly 5 digits
It is necessary that a legal form is allowed in the
country of the business partner’s legal address
2
3
Instantiate
the data to
be validated
as instances
in the CDL
world
Identify and
define the
constraints on
the Corporate
Data League
domain
• Objects: Business Partner, Address, Legal
Form,Country, Aktiengesellschaft, Stock
corporation, Germany, United Kingdom,
Sweden, ….
• Properties: has Address, has Name, has
Definition, has Abbreviation, …
• Literals: «Germany», «A business partner is
…», «67655», …
Business Rules
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 11
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
Business
Partner Address
Country
Germany
An
organization
which has….
Has address
Has country
Is a
A physical
location that…
Has definition
Legal form
Has legal form
Aktiengesel
lschaft
Is a
Has definition
Used in
Rule legal
form
allowed
…
…
Has constraint Rule legal
form valid
…
Has
constraint Rule
country
valid
…
Has
constraint
Name Has name
The CDL data model and business rules are defined in a
graphical user interface and represented in a single ontology
Has legal form
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 12
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
Business partner data is instantiated in the ontology …
CDL 123
(http://corporate-data-
league.ch/CDL_123)
ADDR 123
(http://corporate-data-
league.ch/ADDR_123)
Has country
Business
Partner Address
Country
Germany
Has address
Has country
Is a Legal form
Has legal form
Aktiengesel
lschaft
Is a
Used in
Rule legal
form
allowed
…
…
Rule legal
form valid
…
Is a
Schland
(http://corporate-data-
league.ch/Schland)
Has legal
form
Bayer AG
(http://corporate-data-
league.ch/Bayer_AG)
Has name
Has address
Rule
country
valid
…
Has
constraint Has
constraint
Has
constraint
Name Has name
Is a
Is a
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 13
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
… which is then checked for possible violations of rules in
the « CDL world»
CDL 123
(http://corporate-data-
league.ch/CDL_123)
ADDR 123
(http://corporate-data-
league.ch/ADDR_123)
Has country
Business
Partner Address
Country
Germany
Has address
Has country
Is a Legal form
Has legal form
Aktiengesel
lschaft
Is a
Used in
Rule legal
form
allowed
…
…
Rule legal
form valid
…
Is a
Schland
(http://corporate-data-
league.ch/Schland)
Has legal
form
Bayer AG
(http://corporate-data-
league.ch/Bayer_AG)
Has name
Has address
Rule
country
valid
…
Has
constraint
Has
constraint
Has
constraint
Name Has name
Is a
Is a
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 14
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
It really works!
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 15
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
The approach employs customizable semantic web
technology
Business Partner Data Validator
Business
Rule
Factory
Triple
Factory
Corporate Data League Model
Ontology
External
sources
Ins
tan
ce
s
Ru
les
CDL domain
Validation results
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 16
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
Motivation
Agenda
Corporate Data League
Approach and Application: Semantic Business Rules
Discussion
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 17
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
The approach has major advantages compared to ordinary
solutions (SAP Information Steward, Business rules engines, …)
Plug & Play data validation (batch or realtime): any database can be
easily connected
Make use of mighty inference mechanisms of the ontology (e.g. define what a «German Post Code» is and easily infer which post codes in your database are
German)
Simply add additional, company-specific business rules (including semi-automatic translation of natural language rules)
Provide «Amazon»-like suggestions for fixing data defects or automize
it
Integrate webservices or linked data for data validation and/or data
enrichment
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 18
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
Future opportunities are the development of shared
business rule catalogues
Company-specific
data model &
rules 1
2 Shared business
rule sets
Shared
regulatory
business rule
sets
3
Customizable
dashboard
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 19
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
Explore the Corporate Data League landscape!
https://www.corporate-data-league.ch
Homepage https://www.corporate-data-league.ch/meta
Business vocabulary, rulebook and user/dev guide
https://www.corporate-data-league.ch/app
Demo application
© CDQ AG– St. Gallen, Monat 2015, Schlosser, Simon | University of St. Gallen | 20
65 110 63
151 167 139
164 192 112
222 227 218
222 33 95
238 194 50
www.iwi.unisg.ch
Institute of Information Management
University of St. Gallen
https://www.corporate-data-league.ch
Corporate Data League
www.benchmarking.iwi.unisg.ch
CC CDQ Benchmarking Platform
www.xing.com/net/cdqm
CC CDQ Community at XING
Contact me!
University of St. Gallen
Research assistant
+41 (0)79 9642762
Simon Schlosser