A promising application domain for Semantic Web technology is the annotation of products and services offerings on the Web so that consumers and enterprises can search for suitable suppliers using products and services ontologies. While there has been substantial progress in developing ontologies for types of products and services, namely eClassOWL, this alone does not provide the representational means required for e-commerce on the Semantic Web. Particularly missing is an ontology that allows describing the relationships between (1) Web resources, (2) offerings made by means of those Web resources, (3) legal entities, (4) prices, (5) terms and conditions, and (6) the aforementioned ontologies for products and services. (1NDN)
In the talk, I will explain the need and potential of the GoodRelations ontology, introduce its key conceptual elements, highlight several lessons learned, and summarize design decisions with respect to to modeling approaches and the appropriate language fragment, which may be relevant for other ontology projects, too.
Unleashing Real-time Insights with ClickHouse_ Navigating the Landscape in 20...
The GoodRelations Ontology: Making Semantic Web-based E-Commerce a Reality
1. The GoodRelations Ontology: Making Semantic
Web-based E-Commerce a Reality
http://purl.org/goodrelations/
Martin Hepp
mhepp@computer.org / http://www.heppnetz.de
Martin Hepp
1
mhepp@computer.org
2. E-Commerce on the Web
World Wide Web
Enterprise 1 Enterprise 2
Structured Structured
Data on Data on
Products and Products and
Services Services
(purchased (purchased
and offered) and offered)
Martin Hepp
2
mhepp@computer.org
3. E-Commerce on the Semantic Web
World Wide Web
Enterprise 1 Enterprise 2
Structured Structured
Data on Data on
Products and Products and
Services Services
(purchased (purchased
and offered) and offered)
Product Specifications: Type of Product, Features etc.
Details of the Offering: Price, Eligible Regions, etc.
Martin Hepp
3
mhepp@computer.org
4. 2001-2008:
Semantic Web & E-Commerce:
Lots of Papers, Limited Impact
Martin Hepp
4
mhepp@computer.org
5. Use Cases
• Commodity We sell
some
We sell some
instances of
We sell a
particular
We clean and
repair
offers
sleeping the Marmot used instance sleeping
bags. sleeping bag of the Marmot bags.
model 1234. sleeping bag
• Services
model 1234.
offers
• Product
model data
interchange
Martin Hepp
5
mhepp@computer.org
6. Required Ontologies
• An ontology for product
types and features
• by function, usage, or nature
• An ontology for offer
specifications
• An offer is basically a relation
between
• an agent,
• a set of objects,
• a set of property rights,
• an audience, and
• a set of terms and conditions.
Martin Hepp
6
mhepp@computer.org
7. eClassOWL and GoodRelations
• eClassOWL instanceOf eclassowl:TVSet
foo:myTVSet
• GoodRelations
foo: MillerInc goodrelations:offersToSell foo:myTVSet
Martin Hepp
7
mhepp@computer.org
8. within the domain name space of local dealers.
2.2 Competency Questions
Next, we specify the scope and purpose of the GoodRelations ontology usingcompetency
questions, which is a standard technique in ontology engineering methodologies [8]. The
questions were discussed with stakeholders, namely the developers of recommender
Competency Questions
systems, operators of Web shops, and other domain experts.
CQ1: Which retrievable Web Resources describe an offer
• {to sell | to provide the service of | to repair | to maintain | to lease out | to dis-
pose}
• {a concrete individual | some unknown individuals} of
• a {given good | given service | spare part for a given good | consumables and
supplies for a given good} described by a {type of good | specific make and
model}
• that meet certain requirements on {properties | intervals for properties}
• for which the offering party accepts a given method of payment and
• provides a certain method of delivery
• to {consumers | retailers}
• in a given {country | region}?
CQ2: For which time frame is the Part
+ Upper Ontology offer valid?
CQ3:Hepp
Martin Which types of customers are eligible?
8
mhepp@computer.org
CQ4: Which are the eligible customer regions?
• Overview
CQ5: Which shipping / delivery methods are available?
9. Requirements / Features
• Support for ranges and units of measurements
• Support for all common business functions, like sell, lease, dispose, repair, etc.
• Compatible with eclassOWL and unspscOWL
• Supports all ISO 4217 currencies
• Supports defining eligible regions
• Suits both for explicit instances, product models, and anonymous instances
• Supports common delivery and shipping methods
• Supports accepted payment methods
• Offerings can be constrained to certain eligible business entities
• A warranty promise, i.e., its duration and scope can be specified
• Different prices for different types of customers or for different quantitities can be expressed
• Charges for certain payment or delivery options can be specified; the latter also individually per
region.
• Support for product bundles, for all kinds of units of measurements (2 kg butter + 2 cellphones for €
99 would be no problem).
• Compatible with international standards: ISO 3166, ISO 4217, UN/CEFACT, eCl@ss, and UNSPSC
• Minimal requirements on reasoner support - any RDF-S-style reasoner, OWL DLP, DL, or ter Horst
reasoner will work.
• Supports price ranges, list prices, time-zones, EAN/UCC/UPC codes, GLN/ILN, and DUNS
Martin Hepp
9
mhepp@computer.org
13. Design Considerations
• Suitable Ontology Language
• Ranges and Intervals; Datatypes
• Models, Classes, Instances
• Existential Quantification
• N-Ary Relations
• Licensing
Martin Hepp
13
mhepp@computer.org
14. Suitable Ontology Language
Requirements: Approach: Use OWL DL
syntax for RDF-S elements, i.e. a
• Must work with Semantic
Web infrastructure as subset of the closure of OWL
available today DLP:
•
owl:Ontology
Must work with lightweight, owl:Class
RDFS-style reasoners
owl:ObjectProperty
• Combining the ontology
with ontologies or
owl:DatatypeProperty
knowledge bases in OWL rdfs:subClassOf
DL should not lead to a rdfs:subPropertyOf
model beyond DL (e.g. rdfs:comment
OWL Full)
rdf:datatype
• Avoid layering problems
between RDFS and OWL
rdf:type
Martin Hepp
14
mhepp@computer.org
15. Ranges and Intervals
hasValue rdfs:SubPropertyOf hasMaxValue, hasMinValue
Martin Hepp
15
mhepp@computer.org
16. Models, Classes, Instances, and Existential Quantification
gr:ProductOrService
rdfs:subClassOf rdfs:subClassOf
rdfs:subClassOf rdfs:subClassOf
gr:ActualProductOrServiceInstance gr:ProductOrServicesSomeInstancesPlaceholder
toy:TVSet
rdf:type gr:ProductOrServiceModel
rdf:type
ex:mySony100
rdf:type rdf:type
rdf:type
rdf:type
ex:TheSony100Model Problem: Inferring
Requirement: Use the
feature defaults for product
same library of features for
instances from the product
models and products
ex:SomeSony100s model if specified.
Martin Hepp
16
mhepp@computer.org
18. N-Ary Relationships
ProductOrServiceInstance
typeOfGood
includesObject amountOfThisGood
Offering TypeAndQuantityNode Quantity
hasUnitOfMeasurement
Unit of Measurement
Martin Hepp
18
mhepp@computer.org
19. License
• Goal: Give adopters certainty about
permanent, royalty free access to the ontology
• Approach: Creative Commons Attribution 3.0
license.
Martin Hepp
19
mhepp@computer.org
20. Reusing Existing Consensus
• Lot of pre-existing consensus
• Countries
• Currencies
• Locations
• Two basic approaches for reuse
• As ranges for literal values
• Replication as ontological instances
Martin Hepp
20
mhepp@computer.org
22. Pick-up in Industry and Academia
• Smart Information Systems
• ebSemantics
• Yahoo! SearchMonkey
• Virtuose Sponger Catridges for Amazon, eBay,
and others expected
• etc.
Martin Hepp
22
mhepp@computer.org
23. Diffusion Strategy
• Make Creating GoodRelations
Data Easy
• Make Creating GoodRelations
Data Attractive
• GoodRelations Annotator
and Validator
• Search Engine Vendors
• Exporters for popular Web
• Recommender Systems
Shops (using Triplify) • Foster the Development of
Compatible Vocabularies
• Converters for Catalog Data
Interchange Standards • eClassOWL
• Recipes and Patterns • ebSemantics
• myOntology
Martin Hepp
23
mhepp@computer.org
25. Discussion and Future Extensions
• Richer Axiomatisation?
• Disjointness Axioms etc.
• Microformats Variant?
Martin Hepp
25
mhepp@computer.org
26. Additional Information
• Web Page
•Ontology
•Language Reference
•Primer
•Recipes
•Wiki
http://purl.org/goodrelations/
Martin Hepp
26
mhepp@computer.org
27. References
1. Hepp, Martin: GoodRelations: An Ontology for Describing Products and Services
Offers on the Web, Proceedings of the 16th International Conference on
Knowledge Engineering and Knowledge Management (EKAW2008), Acitrezza,
Italy, September 29 - October 3, 2008 (forthcoming), Springer LNCS, Vol 5268,
pp. 332-347.
2. Hepp, Martin: Products and Services Ontologies: A Methodology for Deriving OWL
Ontologies from Industrial Categorization Standards, in: Int'l Journal on Semantic Web
& Information Systems (IJSWIS),Vol. 2, No. 1, pp. 72-99, January-March 2006.
3. Hepp, Martin: The True Complexity of Product Representation in the Semantic Web, in:
Proceedings of the 14th European Conference on Information System (ECIS
2006), June 12-14, 2006.
4. Hepp, Martin: ProdLight: A Lightweight Ontology for Product Description Based on
Datatype Properties, Proceedings of the 10th International Conference on
Business Information Systems (BIS 2007), April 25-27, 2007, Poznan, Poland, in:
Abramowicz, Witold (Ed.): BIS 2007, Springer LNCS,Vol. 4439, pp. 260-272, 2007.
These and other papers are available at http://www.heppnetz.de/publications/
Martin Hepp
27
mhepp@computer.org
28. Thank you!
http://purl.org/goodrelations/
Martin Hepp
mhepp@computer.org / http://www.heppnetz.de
Martin Hepp
28
mhepp@computer.org