PersoBOX:
A Personalization Engine Between
ERP System and Web Frontend
Dipl.-Inform. Christoph Adolphs
Prof. Dr. Petra Schubert
Research Group Business Software
University of Koblenz-Landau
Department of Computer Science
Institute for IS Research
Agenda
Introduction: What is Personalization?
The PersBOX Project
Relevance
Prior projects
Architectonical overview
Future research
What is Personalization?
Personalization is …
“about building customer loyalty by building
meaningful one-to-one relationships; by understanding
the needs of each individual and helping satisfy a goal
that efficiently and knowledgeably addresses each
individual’s need in a given context.”
Riecken, 2000
”the adjustment and modification of all aspects of a
website that are displayed to a user in order to match
that users needs and wants.”
Additional definition of Personalization
… the individual adaptation of content and
functionalities of (e-commerce) applications to the
needs of a user. The adaptation is based on implicitly
or explicitly received and stored user data.
according to Risch, 2007
According to prior projects we define
R
e
la
tio
n
In
te
ra
c
tio
n
Personalization Framework
S
o
u
rc
e
:
R
is
c
h
2
0
0
7
-F
o
llo
w
in
g
V
e
s
a
n
e
n
2
0
0
5
Personalized Marketing
Output
Promotion / Communication
Price
Place / Delivery
Product / Services
Customer
tra
ns
ac
tio
n
Value for
customer
Cost
for customer
Benefit for
customer
Better preference
match
(Better
products)
Better
service
Better
Communication
Experience
of one
Privacy
risks
Spam
risks
Spent
time
Extra
fees
(Waiting
costs)
Cost
for business
Benefit for
business
Customer
loyalty
Higher Prices
Differentiation
Higher Turnover
(Cross-/Up-Selling)
Better response
rates
Risk of irritating
customers
Channel
conflicts
Risk of loosing
trust
Investments in
technology
Investments
in education
Business
Value for
business
Company with
ERP system
data
transmission
master data
adminstration
PersoBOX operator
basic settings
start
personalization
process
data
send data as
personalized
functions
(combine data
+ functions)
update
Data Store
E-Shop operator
page request
page creation
data
transmission
master data
administration
page assembly
Customer
page visit
page display
select personalized
info
timeline
customizing
customizing
Customer Profile Life Cycle
Learning from user behaviour / interaction
Input for Redesign
Plan/Model
• Requirements
/ Availability
• Source
• Structure
• Storage
Input Profile
• Identification
• Preferences
• Interaction
• Transaction
• Context
• Ratings
Methods and
Techniques
• Data Mining
• OLAP
• Web
Analytics
• Rule
Engines
Output Profile
• Customer Value
• Priority
•
Recommen-dations
• Clusters
• Classifications
Usage and
Application
„Output“
Application
• Personalization
• Customization
• Segmentation
• Marketing
Campaigns
• Documentation
• Selling
Gathering
• explicit
• implicit
Integration
• ETL
• Data
Warehouse
Analysis and
Processing
„Processing“
Gathering
and
Integration
„Input“
Planning and
Modelling
Customer Profiles and Personalization
Processing
Web Logfile
ERP
Transactions
ERP
Products
ERP
Cust. Profiles
& Conditions
CRM
External
Data
Data Warehouse
Meta-Data
Meta-Data
Meta-Data
User
Profile
Product
Data
Page
Content
ETL (Input Profiles)
Rule Engine
Automated
Rules
Data Marts containing individual profiles and content (Output Profiles)
E-Shop
1
2
3
4
5
6
7
User
Settings &
Preferences
8
Other Applications /
Services
E-Shop DB
Output Profile
(Information)
Input Profile
(Data)
S
o
u
rc
e
:
a
c
c
o
rd
in
g
t
o
R
is
c
h
/
S
c
h
u
b
e
rt
/
L
e
im
s
to
ll
2
0
0
6
,
p
.
5
Data processing for personalization purposes
Defining input Interfaces (Customizing)
Input processing
Unifying
Filtering
Storing
Output processing
Generating and using of customer profiles
Generating function instances
Applying data to instances
Architecture of PersoBOX
Learning from user behaviour / interaction
Input for Redesign
Usage and
Application
„Output“
Analysis and
Processing
„Processing“
Gathering
and
Integration
„Input“
Planning and
Modelling
Filtering rules
Input profile
3rd Party System
Data 3rd Party
System
Input Schema
Data web shop
Input Schema
Data ERP System
G
Input
Input processing
Output processing
Output
System
Input Profile
A
Output forecast:
e.g. platform,functions, design or callback functions
B
ru
le
b
a
s
e
d
f
il
te
ri
n
g
:
e
lim
in
a
ti
n
g
u
s
e
le
s
s
d
a
ta
b
a
s
e
d
o
n
t
h
e
o
u
tp
u
t
fo
re
c
a
s
t
C
p
a
tt
e
rn
b
a
s
e
d
u
n
if
ic
a
ti
o
n
:
S
a
v
in
g
o
f
d
a
ta
i
n
s
tr
u
c
tu
re
s
,
b
a
s
e
d
u
p
o
n
p
a
tt
e
rn
s
f
o
u
n
d
i
n
a
m
a
tc
h
in
g
d
a
ta
b
a
s
e
unifying patterns
N
E
Output profile
personalized set
of applications
F
Preference Profile Builder:
Influence on user preferences based
on input profile analysis
H
Function chooser:
Based upon the output forecast and the
analysation of the user preferences a set of
personalisation functions can be loaded with
reference to the unified data
J
Dynamic Personalized
Application Generator:
Dynamic implementation dependent
on user preferences an output
forecast
M
Deployment process:
Static or dynamic deployment by using
different techniques (e.g. as applett)
K
Data transfer Scheduler:
Transfers relevant data
into a working database for
not corrupting existing data
Set of
personalized
user data
Set of
personalisation
functions
I
L
E.1
E.2
F.1
F.2
G.3
F.4
G.2
K.1
L.1
M.2
M.1
J.2
J.1
H.1
H.4
H.2
H.3
A.4
A.1
A.2
A.3
C.1
C.2
C.3
B.2
B.3
B.4
B.1
N.1
unified data
based upon
forecast
J.3
D
C.4
D.1
G.1
E.3
C
us
to
m
er
pro
file
D
is
pl
ay
u
se
r
pr
of
ile
O
Request scheduler
O.1
I.1
I.2
Request
customizing profile
A.5
O.2
P
Data Input Scheduler:
Triggers data transmission
internal or external events
Q
C.5
Future Research
Creating a fine planned architecture
Identifying potential project partners
Implementing a prototype fulfilling different aspects
of the PersoBOX architecture
Taxonomies for filtering or unifying the data
Automatic code generation
Intelligent function choosing
Thank you for your attention.
Dipl.-Inform. Christoph Adolphs
Prof. Dr. Petra Schubert
Research Group Business Software
University of Koblenz-Landau
Department of Computer Science
Institute for IS Research
Literature
Riecken, Doug (2000): Personalized Views of Personalization, in: Communications of
the ACM, Volume 43, No. 8, 2000.
Risch, Daniel (2007): Kundenprofile im E-Commerce - Ergebnisse einer empirischen
Studie zum Umgang mit Kundendaten im Electronic Commerce, Arbeitsbericht
E-Business Nr. 29, Basel: Fachhochschule Nordwestschweiz - Institut für
Wirtschaftsinformatik, 2007.
Schubert, Petra; Kummer, Mathias; Leimstoll, Uwe (2006):Legal Requirements for the
Personalization of Commercial Internet Applications in Europe, in: Journal of
Organizational Computing and Electronic Commerce 16 (3&4), 203–220, 2006.
Vesanen, Jari (2005): What is Personalization? –
A Literature Review and Framework, Helsinki: Working Paper, Helsinki School of
Economics, 2005.
Wu, Dezhi; Im, Il; Tremaine, Marilyn; Instone, Keith; Turoff, Murray (2003): A
Framework for Classifying Personalization Scheme Used on e-Commerce Websites,
in: Proceedings of the 36th Hawaii International Conference on System Sciences,
HICSS’03, Hawaii, 2003.
Risch, Daniel ; Schubert, Petra ; Leimstoll, Uwe (2006): “The Personalization Map –
An Application-Oriented Overview of Personalization Functions.” In: Proceedings of
the Joint Conference of the International Mass Customization Meeting (IMCM’06) and
the International Conference on Economic, Technical and Organizational Aspects of
Product Configuration Systems (PETO’06). Hamburg, 2006
Customers influence on personalization process
output
Customer
profile
request
output transition
d
a
ta
st
o
re
predicted
request
consumer
e-shop
ERP system
transaction data
user profiles
CRM data
ERP data
data and input transition
data and input transition
output deployment
explicit influence
implicit influence
legend:
Customer Profile
Product Profile
Organization of Product Database
ProdCat {sports, events, garments,
shoes, electronics, food, …}
PriceCat {low, middle, high}
ProdGroup SPORTS {tennis, golf,
joggin, ski, trekking, …}
ProdGroup EVENTS {region, type, …}
Marketing Rules
1. Event.Basel.HighPrice + Sports.Tennis
Tickets Swiss Indoors Basel
2. Sports.Tennis + High turnover for
ProdGroup Sports.Tennis
New Nike indoor tennis shoes
3. Purchased Electronics.DVDs.Fantasy
New Harry Potter DVD
Web site (E-Shop)
Registration
Interests {tennis, golf, DVDs, …},
age, region {Basel, Zurich, …}
ClickStream
ProdCat, ProdGroup, …
Customer value card
Shopping transactions
Date, ProdCat, ProdGroup,
PriceCat, …
Marketing measures
Customer reaction towards
offers and discounts e.g.
event (region, type, ZIP, …)
IN
P
U
T
P
ro
fi
le
Web site (E-Shop)
Registration
Interests {tennis, golf, DVDs, …},
age, region {Basel, Zurich, …}
ClickStream
ProdCat, ProdGroup, …
Customer value card
Shopping transactions
Date, ProdCat, ProdGroup,
PriceCat, …
Marketing measures
Customer reaction towards
offers and discounts e.g.
event (region, type, ZIP, …)
Web site (E-Shop)
Registration
Interests {tennis, golf, DVDs, …},
age, region {Basel, Zurich, …}
ClickStream
ProdCat, ProdGroup, …
Customer value card
Shopping transactions
Date, ProdCat, ProdGroup,
PriceCat, …
Marketing measures
Customer reaction towards
offers and discounts e.g.
event (region, type, ZIP, …)
IN
P
U
T
P
ro
fi
le
Sports.Tennis, Sports.Golf, Sports.Ski,
Events.Basel.Highprice,
Electronic.DVDs.Fantasy,
Electronic.DVDs.ScienceFiction,
Turnover.Sports.Tennis=high
T
P
U
T
P
ro
fi
le
Sports.Tennis, Sports.Golf, Sports.Ski,
Events.Basel.Highprice,
Electronic.DVDs.Fantasy,
Electronic.DVDs.ScienceFiction,
Turnover.Sports.Tennis=high
T
P
U
T
P
ro
fi
le
1. Tickets 27.10.2005
Swiss Indoors
Basel
2. New Nike indoor tennis shoes
Deduction of
customer attributes
Application of
rules on products
e
:
a
c
c
o
rd
in
g
t
o
R
is
c
h
2
0
0
7
The Personalization Map
An Application-oriented
Overview on Personalization
Features
Complementary Activities 5 Screen design5.1 Welcome, addressing the user 5.2 Design options 5.2.1 MyHomepage 5.2.2 Colors 5.2.3 Forms 5.2.4 Portlets 5.3 Menu (horizontal) 5.4 Menu (vertical) 6 Community 6.1 Reviews 6.2 Hit lists 6.3 Collaborative Filtering 6.3.1 Basic functions 6.3.2 Soul sisters 6.3.3 Collaborative categories 6.4 Ratings 6.4.1 Product ratings 6.4.2 Vendor ratings 6.4.3 Rating of reviews 7 Customer profile
and role concept (User accounts)
7.1 Role concept
7.1.1 Individual user profile (single user) 7.1.2 Role-based profile (multiple users) 7.1.3 Administrator 7.2 Access to
customer data 7.2.1 View profile 7.2.2 Change profile
8 Marketing and CRM
8.1 Newsletter,
hint and news 8.1.1 Information history-based 8.1.2 Cross-selling history-based 8.2 Alerts (Scheduler)
8.2.1 Event-triggered news 8.2.2 Advertisement 8.2.3 Offers
8.2.4 Reminders (e.g. renewal date) 8.3 Cross- und Up-Selling
8.4 Advertisements
8.4.1 Banners 8.4.2 Interstitials 8.4.3 Pop-ups 8.4.4 Cross-links, web rings 8.4.5 E-cards 8.5 Entertainment
8.5.1 Competitions 8.5.2 Games 8.5.3 Video files
9 Reports and Statistics
9.1 Order process
9.1.1 Order history 9.1.2 Analysis 9.1.3 Individual top sellers 9.2 Click stram analysis
9.3 Data Mining Buying Process 1 Information Phase 1.1 Electronic product catalog (EPC) 1.1.1 Search 1.1.2 Product structure 1.1.3 Customer-specific assortment 1.1.4 Personal shopping lists 1.1.5 Sample shopping lists 1.1.6 Compatibility lists 1.1.7 Translator (substitution lists)
1.2 Product presentation 1.2.1 Sales promotion page
1.2.2 Recommendations 1.2.3 Product/Service bundle 1.2.4 Service contract 1.2.5 Cross-/Up-Selling 1.2.6 Gift ideas 1.2.7 Product novelties 1.3 Pricing and Conditions 1.3.1 Individual prices 1.3.2 Special prices 1.3.3 Discount programs 1.3.4 Bonus programs 1.3.5 Bundling 2 Agreement Phase 2.1 Configuration of products and services 2.1.1 Mass customization 2.1.2 Product configurators 2.1.3 Integration of third party configurators
2.2 Calculation of prices based on product configuration
2.3 Request for quotation 2.4 Negotiation of conditions 2.5 Shopping cart 2.6 Check-out support 2.6.1 Delivery options 2.6.2 Payment options 3 Settlement Phase 3.1 Automatic order placement
3.1.1 Regular amount and frequency 3.1.2 Suggestions for ideal amount
3.2 Automatic delivery trigger 3.2.1 Subscription
3.2.2 Automatic replenishment
3.3 Tracking und Tracing
4 Supporting Functions 4.1 Order process guidance
4.1.1 Wizard 4.1.2 Avatar 4.1.3 Personal consultant 4.1.4 Call center 4.1.5 Co-browsing 4.2 Special B2B functions 4.2.1 Support for intermediaries
4.2.2 Storing of matching article numbers 4.2.3 Implementation of approval process
