Challenges and Trends for Personalization in the Semantic Web

pikeactuaryInternet and Web Development

Oct 20, 2013 (3 years and 9 months ago)

58 views

Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Challenges and Trends for
Personalization in the Semantic Web
Nicola Henze
ISI - Semantic Web Group &
Research Center L3S
University of Hannover
henze@l3s.de
PerSWeb’05
Workshop on Personalization on the Semantic Web
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Outline
1
Introduction
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
2
Personalization Revisited
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
3
Challenges for Personalization in the Semantic Web
Machine Processable Semantics
Service-oriented Technologies
4
Trends
Some Recent Approaches
My Conclusions
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
Outline
1
Introduction
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
2
Personalization Revisited
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
3
Challenges for Personalization in the Semantic Web
Machine Processable Semantics
Service-oriented Technologies
4
Trends
Some Recent Approaches
My Conclusions
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
Office Hours at a University Department
A User Question:
“Can I come to the office
on July 25th at 6 p.m.?”
Required Processing:
which counseling
schedule is to apply?
→consult Web page
which weekday is July
25th?
→consult a calendar
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
Office Hours at a University Department
A User Question:
“Can I come to the office
on July 25th at 6 p.m.?”
Required Processing:
which counseling
schedule is to apply?
→consult Web page
which weekday is July
25th?
→consult a calendar
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
Office Hours at a University Department
A User Question:
“Can I come to the office
on July 25th at 6 p.m.?”
Required Processing:
which counseling
schedule is to apply?
→consult Web page
which weekday is July
25th?
→consult a calendar
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
Analyzing the Scenario
A User Question:
“Can
I
come to the office on July 25th at 6 p.m.?”
Easy for humans
at least:time consuming
non-ambiguous data:office hours,time schedule,calendar
data
Suitable for machine processing!
Personalization:
localization,contextualization,preferences,my time
schedule,etc.
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
Analyzing the Scenario
A User Question:
“Can
I
come to the office on July 25th at 6 p.m.?”
Easy for humans
at least:time consuming
non-ambiguous data:office hours,time schedule,calendar
data
Suitable for machine processing!
Personalization:
localization,contextualization,preferences,my time
schedule,etc.
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
And this is what a machine “understands”
Missing:
Machine-processable
meaning of data
interpretation
office hour is a kind of
opening hour,which
has the meaning...
reasoning
IF July 25th is a
Monday AND office
hours on Mondays..
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
And this is what a machine “understands”
Missing:
Machine-processable
meaning of data
interpretation
office hour is a kind of
opening hour,which
has the meaning...
reasoning
IF July 25th is a
Monday AND office
hours on Mondays..
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
And this is what a machine “understands”
Missing:
Machine-processable
meaning of data
interpretation
office hour is a kind of
opening hour,which
has the meaning...
reasoning
IF July 25th is a
Monday AND office
hours on Mondays..
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
And this is what a machine “understands”
Missing:
Machine-processable
meaning of data
interpretation
office hour is a kind of
opening hour,which
has the meaning...
reasoning
IF July 25th is a
Monday AND office
hours on Mondays..
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
Outline
1
Introduction
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
2
Personalization Revisited
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
3
Challenges for Personalization in the Semantic Web
Machine Processable Semantics
Service-oriented Technologies
4
Trends
Some Recent Approaches
My Conclusions
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
A Web of Semantics
The Semantic Web Vision
"The Semantic Web is an extension of the current web in which
information is given well-defined meaning,better enabling
computers and people to work in cooperation".
TimBerners-Lee,James Hendler,Ora Lassila.Scientific American,May 2001
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
Outline
1
Introduction
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
2
Personalization Revisited
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
3
Challenges for Personalization in the Semantic Web
Machine Processable Semantics
Service-oriented Technologies
4
Trends
Some Recent Approaches
My Conclusions
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
Semantic Web Architecture
Semantic Web Architecture
1.Languages
for describing Web resources and their
relations(RDF)
for modeling contexts,domains,
interpretations (Ontology Languages,OWL)
for annotating Web resources with rules &
policies (Rule ML,SWRL,...)
Remember scenario:Machine-processable
interpretation of data!
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
Semantic Web Architecture
Semantic Web Architecture
1.Languages
2.Reasoning
rule-based languages for reasoning on
Resource descriptions & ontological
knowledge
non-monotonic reasoning,defeasible rules,
etc.
Remember scenario:Reasoning on
machine-processable meaning of data!
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
Semantic Web Architecture
Semantic Web Architecture
1.Languages
2.Reasoning
3.Proof & Trust
verification,negotiation,security
Remember scenario:Trust the concluded facts;
trust the information provider;negotiate level of
understanding,...
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
Semantic Web Architecture
Figure:
The Semantic Web Protocol Stack as of Sep.2002
http://www.w3.org/2002/Talks/09-lcs-sweb-tbl/
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
Outline
1
Introduction
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
2
Personalization Revisited
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
3
Challenges for Personalization in the Semantic Web
Machine Processable Semantics
Service-oriented Technologies
4
Trends
Some Recent Approaches
My Conclusions
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
Personalization for Information Systems
Information is filtered according to a user’s particulars
(context,preferences,goal,etc.).
The filter can be learned (“adaptive”) and/or can be
adjusted by the user (“adaptable”).
Adaptive and adaptable approaches can be mixed.
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
Example (Pull Scenario:The user requests information)
information request is translated into a query
narrow search:e.g.query refinement,query rewriting
information is retrieved
rate content:e.g.ranking of results
information is selected
finalize result:device constraints,costs,quality,..
information is shipped
user interface:presentation,create awareness of the
personalization process
Filtering can take place during all phases!
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
Example (Push Scenario:The system continuously filters data)
A filter can be employed to pro-actively recommend information
to the user
which fits to the user’s particulars
which fits to observations about other users in similar
contexts
which relates to currently browsed resources
which points out information “close-by”
etc.
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
Outline
1
Introduction
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
2
Personalization Revisited
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
3
Challenges for Personalization in the Semantic Web
Machine Processable Semantics
Service-oriented Technologies
4
Trends
Some Recent Approaches
My Conclusions
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
Web Mining-based Personalization
Structural analysis,pattern
discovery
Content:
similar/related
content (“edges”)
Structure:
content quality,
content close-by (“vertices”)
Usage:
content used under
similar conditions
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
Web Mining-based Personalization:Benefits & Drawbacks
+
well-known statistical methods for analysis of
structure,recognition of patterns,etc.
+
minimalistic user profiling required
+
success especially in e-commerce applications
-
granularity of recommendations is limited
-
slow to adapt to new contexts
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
Outline
1
Introduction
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
2
Personalization Revisited
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
3
Challenges for Personalization in the Semantic Web
Machine Processable Semantics
Service-oriented Technologies
4
Trends
Some Recent Approaches
My Conclusions
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
Hyperspace
Graph:
vertices:
hypermedia
documents
edges:
pairs of oppositely
directed hyperlinks
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
Personalized Hyperspace
Constructed dynamically according
to a users needs
vertices:
selection of content/
parts of content
content adaptation
edges:
selection/annotation
/grouping/creation
of hyperlinks
navigational
adaptation
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
Adaptive Hypermedia requires:
Metadata on vertices:
keywords for content,parts of
content,role of content,etc.
Relations between vertices:
“is_prerequisite”,“requires”,
“is_alternative_explanation”,
“deepens”,“gives_details”,
etc.
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
Adaptive Hypermedia:Benefits & Drawbacks
+
success on well-defined,relatively small spaces
+
provide guidance & orientation
+
success especially in e-Learning applications
-
closed space
-
lack of re-usability
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Machine Processable Semantics
Service-oriented Technologies
Outline
1
Introduction
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
2
Personalization Revisited
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
3
Challenges for Personalization in the Semantic Web
Machine Processable Semantics
Service-oriented Technologies
4
Trends
Some Recent Approaches
My Conclusions
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Machine Processable Semantics
Service-oriented Technologies
Semantic Web Challenges & Benefits
Machine processable Semantics
=⇒
additional (reliable) information for analyzing the
Web Graph
mine Resource descriptions (vertices &
edges)
mine usage
=⇒
additional (reliable) information on Web Resources
and their relations
adaptively select (parts of) Resources
adaptively identify relations of interest,create
shortcuts,...
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Machine Processable Semantics
Service-oriented Technologies
Outline
1
Introduction
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
2
Personalization Revisited
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
3
Challenges for Personalization in the Semantic Web
Machine Processable Semantics
Service-oriented Technologies
4
Trends
Some Recent Approaches
My Conclusions
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Machine Processable Semantics
Service-oriented Technologies
Service-oriented Technologies
(Semantic) Web Services
Web Services:
self-contained,self-describing,modular
applications that can be puslihed,located,and
invoked accross the Web.
Semantic Web Services:
combining Web Services technology
with Semantic Web technology (automated
discovery,combination,execution of Web
Services)
We can distinguish:
Services offering Personalization
Personalization of Services
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Some Recent Approaches
My Conclusions
Outline
1
Introduction
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
2
Personalization Revisited
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
3
Challenges for Personalization in the Semantic Web
Machine Processable Semantics
Service-oriented Technologies
4
Trends
Some Recent Approaches
My Conclusions
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Some Recent Approaches
My Conclusions
Personalized Content Syndication
Example (A syndicated view on a research project)
publications are success indicators for research =⇒online!
research projects maintain information portals =⇒online!
gather this online information,integrate heterogenous
sources,syndicate according to user’s preferences
Realization:
semi-automated extraction of publication information from
heterogenous sources
personalization rules reason about publication data and
project data,e.g.to recommend related publications or
research
R.Baumgartner,N.Henze,and M.Herzog,ESWC 2005.
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Some Recent Approaches
My Conclusions
Context-Aware Personalization in e-Commerce
Example (Route-planning offered across applications)
An e-commerce application offering information to tourists:
offers information about tourist attractions + information on
how to get there (route planning),
later on,a hotel booking services is added to the
e-commerce application,
to enhance user satisfaction,a route planning should be
offered by this new service,too.
Realization:
context information is pre- and post-processed by
dedicated context plug-ins/context services
extension of SOAP protocol
M.Keidl and A.Kemper,WWW2004.
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Some Recent Approaches
My Conclusions
Personalized e-Learning in the Semantic Web
Example (Offering individual learning support across courses)
1
embed Learning Objects with a personalized context
2
enable learners to choose which kind of personalized
guidance in what combination they appreciate as support
(plug & learn)
Realization:
re-usable personalization algorithms reason about
distributed data sources (user data,course descriptions,
ontologies,etc.)
personalization algorithms are exposed as Web Services
N.Henze,AIED 2005
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Some Recent Approaches
My Conclusions
Outline
1
Introduction
A Familiar Scenario
The Semantic Web Vision
Architecture of the Semantic Web
2
Personalization Revisited
Personalized Information Systems
Web Mining-based Personalization
Adaptive Hypermedia
3
Challenges for Personalization in the Semantic Web
Machine Processable Semantics
Service-oriented Technologies
4
Trends
Some Recent Approaches
My Conclusions
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Introduction
Personalization Revisited
Challenges for Personalization in the Semantic Web
Trends
Some Recent Approaches
My Conclusions
My Conclusions
As the Semantic Web Architecture emphasizes rules:
Think about rule-based personalization!
Think about personalization rules on resources!
Think about rule-based control & syndication!
As Semantic Web deals with heterogenous,dynamic
information:
Think about re-usable personalization functionality!
Think about offering your personalization functionality or
your system as services!
Register your Semantic Web Personalization Services!
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Appendix
References
References
I
Rules & Reasoning on the Semantic Web.
The Network of Excellence “REWERSE - Reasoning on the
Web”.www.rewerse.net.
The Personal Reader Framework for designing,
implementing & maintaining Personalization Services
www.personal-reader.de.
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web
Appendix
References
References
II
Markus Keidl and Alfons Kemper
Towards Context-Aware Adaptable Web Services.WWW
2004.
Robert Baumgartner,Nicola Henze,and Marcus Herzog.
The Personal Publication Reader:Illustrating Web Data
Extraction,Personalization and Reasoning for the
Semantic Web.ESWC 2005.
Nicola Henze
Challenges and Trends for Personalization in the Semantic Web