Page 1
The Web of Data for E-Commerce in Brief
A Hands-on Introduction to the GoodRelations Ontology, RDFa, and Yahoo! SearchMonkey
October 25, 2009
Westfields Conference Center near Washington, DC, USA
Martin Hepp Universität der Bundeswehr München, Munich, Germany
Richard Cyganiak Digital Enterprise Research Institute (DERI), Ireland
Page 2
About the Organizers
Martin Hepp
Professor, Head of Group
Universität der Bundeswehr München
Munich, Germany
[email protected]
http://www.heppnetz.de
Previous affiliations: Universität
Würzburg (Germany), Florida Gulf
Coast University, IBM Zurich Research Lab, DERI/STI
Innsbruck
Richard Cyganiak
PhD Researcher
Digital Enterprise Research Institute
(DERI), Galway, Ireland
[email protected]
http://www.deri.ie
Previous affiliations: FU Berlin,
Germany
25.10.2009 2
Page 3
Learning Goals
Participants will learn
• to use
– the GoodRelations conceptual structures and
– the RDFa syntax
to augment static and dynamic Web sites by the various relevant
details of a commercial Web presence;
• RDFa modeling patterns for more complex RDF structures;
• to publish data on the Semantic Web and make it available for
indexing services, repositories, Yahoo SearchMonkey and
applications;
• to query the Web of Data using SPARQL, and
• the development of simple Yahoo SearchMonkey and Yahoo
BOSS applications.
25.10.2009 3
Page 4
08:30-10:30 Overview and Motivation: Why the Web of Data is Now 30’
Quick Review of Prerequisites 15’ The GoodRelations Ontology: E-Commerce on the Web of Data 75’
10:30-10:45 Coffee Break
10:45-12:30 RDFa: Bridging the Web of Documents with the Web of Data 45’
Expressing GoodRelations in RDFa: A Running Example 30’ GoodRelations – Advanced Topics 30’
12:30-13:30 Lunch Break
13:30-16:00 Hands-on Exercise: Annotating a Web Shop 60’
Querying the Web of Data for Offerings – SPARQL 15’ Querying the Web of Data – Exercises 15’
16:00-16:30 Coffee Break
16:30-18:00 Publishing Semantic Web Data: Make Your RDF Available 30’
Yahoo SearchMonkey and Yahoo BOSS 45’ Discussion, Conclusion, Feedback Round 15’
4
Logistics
Page 5
Resources: Information • Wiki page
http://tr.im/srGx http://www.ebusiness-unibw.org/wiki/Web_of_Data_for_E-Commerce_Tutorial_ISWC2009
• GoodRelations Primer http://www.heppnetz.de/projects/goodrelations/primer/
• GoodRelations Documentation http://purl.org/goodrelations/v1
• RDFa http://www.w3.org/TR/2008/REC-rdfa-syntax-20081014/
• SPARQL http://www.w3.org/TR/rdf-sparql-query/
• Yahoo SearchMonkey http://developer.yahoo.com/searchmonkey/smguide/
25.10.2009 5
Page 6
Resources: Tools
• RDF Validator (and Visualizer) http://www.w3.org/RDF/Validator/
• GoodRelations Annotator http://www.ebusiness-unibw.org/tools/goodrelations-annotator/
• PyRDFa http://www.w3.org/2007/08/pyRdfa/
• Twinkle http://www.ldodds.com/projects/twinkle/
Custom configuration file from Wiki
• RDF2dataRSS http://www.ebusiness-unibw.org/tools/rdf2datarss/
25.10.2009 6
Page 7
Overview and Motivation: Why the
Web of Data is Now
Martin Hepp
25.10.2009 7
Page 8
Limitations of the Web, 2009
Page 9
Specificity vs. Keyword-based Search
• Synonyms
• Homonyms
• Multiple languages
• No parametric
search
9
Page 10
No Unified View: Jumping Back and Forth
Across Data Silos
10
Site
1
Site
2
Site
3
Page
1
Page
2
Page
3
Page
5
Page
7
Page
6
Page
8
Page
4
Searc
h E
ngin
e R
esults
Searc
h E
ngin
e R
esults
Searc
h E
ngin
e R
esults
Searc
h E
ng
ine R
esu
lts
Page 11
We know the best hits only when done.
11
Site
1
Site
2
Site
3
Page
1
Page
2
Page
3
Searc
h E
ngin
e R
esults
Page
5
Page
7
Page
6
Page
8
Page
4
Page 12
Limited Ability to Reuse Data
12
Page 13
The Web: A Bottleneck for Sharing
Product Data
13
Page 14
Web of Data (“Semantic Web”)
14
Page 15
E-Commerce on the Web of Data
15
Page 16
Goal: A Unified View on Commerce
Data on the Web
16
Product Model
Master Data Shop
Offerings Auctions Spare Parts &
Consumables
Warranty
Delivery Payment
Retailers Manufacturers
Arbitrary Query
Extraction
and Reuse
Page 17
On the Shoulders of Giants
17
A Unified View of Commerce Data
on the Web Martin Hepp,
mhepp@compu
ter org
Page 18
Deep Comparison Shopping
18 Martin Hepp,
mhepp@compu
ter org
Site
1
Site
2
Site
3
Page
1
Page
2
Page
3
Page
5
Page
7
Page
6
Page
8
Page
4
Search Engine Results
Page 19
Use Case 1: Product Search
• Find all MP3 players
that have a USB
interface and a color
display, and sort them
by weight (lightest
first).
...on a Web Scale!
19
Page 20
Use Case 2: Product Model Data Reuse
20
Manufacturer
Structured Data on
Products and Services
Retailer / Web Shop
Structured Data on Products
and Services
World Wide WebWorld Wide Web
Product Specifications: Type of Product, Features etc.
Page 21
Use Case 3: Fine-grained Affiliate
Marketing
21
Offers of
computer
add-ons that have
an USB
interface
Screenshot from http://en.wikipedia.org/wiki/USB
Page 22
The Web of Linked Data, Essentially: 1. Cluster Web links by what they mean
2. Use URIs to indicate the type of links
3. Use HTTP URIs so that it is quick and easy to explore
what this URI means.
4. Make clear whether you are referring to something or
its representation.
22 Martin Hepp,
mhepp@compu
ter org
Page 23
The Web of Linked Data, Essentially: 1. Cluster Web links by what they mean
2. Use URIs to indicate the type of links
3. Use HTTP URIs so that it is quick and easy to explore
what this URI means.
4. Make clear whether you are referring to something or
its representation.
23 Martin Hepp,
mhepp@compu
ter org
Page 24
The Web of Linked Data, Essentially: 1. Cluster Web links by what they mean
2. Use URIs to indicate the type of links
3. Use HTTP URIs so that it is quick and easy to explore
what this URI means.
4. Make clear whether you are referring to something or
its representation.
24 Martin Hepp,
mhepp@compu
ter org
Page 25
Technical Effects & Working Assumption
• This will reduce the
computational
complexity of
processing,
combining, reusing
data on a Web scale
25 Martin Hepp,
mhepp@compu
ter org
Page 26
Both Sides Can Help Build a Bridge
26 Martin Hepp,
mhepp@compu
ter org
Page 27
The Web of Linked Data is NOW and HERE
• RDFa has become a W3C Recommendation
– HTML5+RDFa Specification well underway, too
• Yahoo SearchMonkey and BOSS
• Google adopts RDFa
• GoodRelations ontology
• SPARQL Query language and endpoint interface
• Scalable, commercial repositories
• Linked Data Guidelines: Best Practices for co-
existence of the Web of Data and existing Web content
25.10.2009 27
Page 28
NOW and HERE: Yahoo & GoodRelations
25.10.2009 28
Page 29
NOW and HERE: Google (Mock-up)
25.10.2009 29
Page 30
NOW and HERE: OpenLink Virtuoso Spongers
25.10.2009 30
Page 31
GoodRelations #2 of all Web Ontologies
25.10.2009 31
…and this does not yet include the > 10 Mio. offers
from Amazon and eBay!
Page 32
GoodRelations #2 of all Web Ontologies
25.10.2009 32
Page 33
NOW and HERE: BestBuy
• Details on all 1000+ stores in the US using
GoodRelations
– http://stores.bestbuy.com/sitemap.xml
– http://lod.openlinksw.com/sparql
• Full Catalog: >432,000 item descriptions
– http://products.semweb.bestbuy.com/sitemap.xml
– updated on a daily basis
25.10.2009 33
Page 34
Thank you.
25.10.2009 34