Title: Metadata
1Metadata for Web-based Information Management
Dickson K. W. CHIU Senior Member, IEEE ACM Dickson Computer Systems Hong Kong kwchiu_at_acm.org, dicksonchiu_at_ieee.org Poon, Joe Kit Man Lam, Wai Chun Tse, Chi Yung Sui, William Hi Tai Poon, Wing Sze Department of Computer Science, University of Hong Kong
2Towards a Semantic Web
- WWW is an impressive success
- amount of available information (gt 1 Giga-page)
- number of human users (gt 200 Mega-user)
- The current Web represents information using
- natural language (English, Hungarian, Chinese,)
- graphics, multimedia, page layout
- Humans can process this easily
- can deduce facts from partial information
- can create mental associations
- are used to various sensory information
- (well, sort of people with disabilities may have
serious problems on the Web with rich media!)
3Need for understanding Web info
- Tasks often require to combine data on the Web
- hotel and travel infos may come from different
sites - searches in different digital libraries
- etc.
- Again, humans combine these information easily
- even if different terminologies are used!
4What is the Problem?
- Markup comprise
- rendering information (e.g., font size and
colour) - Hyper-links to related content
- Semantic content is accessible to humans but not
(easily) to computers
Consider a typical web page
5What information can we see
- WWW2002
- The eleventh international world wide web
conference - Sheraton waikiki hotel
- Honolulu, hawaii, USA
- 7-11 may 2002
- 1 location 5 days learn interact
- Registered participants coming from
- australia, canada, chile denmark, france,
germany, ghana, hong kong, india, ireland, italy,
japan, malta, new zealand, the netherlands,
norway, singapore, switzerland, the united
kingdom, the united states, vietnam, zaire - Register now
- On the 7th May Honolulu will provide the backdrop
of the eleventh international world wide web
conference. This prestigious event - Speakers confirmed
- Tim berners-lee
- Tim is the well known inventor of the Web,
- Ian Foster
- Ian is the pioneer of the Grid, the next
generation internet
6Information a machine may see
- WWW2002
- The eleventh international world wide web
conference - Sheraton waikiki hotel
- Honolulu, hawaii, USA
- 7-11 may 2002
- 1 location 5 days learn interact
- Registered participants coming from
- australia, canada, chile denmark, france,
germany, ghana, hong kong, india, ireland, italy,
japan, malta, new zealand, the netherlands,
norway, singapore, switzerland, the united
kingdom, the united states, vietnam, zaire - Register now
- On the 7th May Honolulu will provide the backdrop
of the eleventh international world wide web
conference. This prestigious event - Speakers confirmed
- Tim berners-lee
- Tim is the well known inventor of the Web,
- Ian Foster
- Ian is the pioneer of the Grid, the next
generation internet
7Solution XML markup with meaningful tags?
ltnamegtWWW2002 The eleventh international world
wide webconlt/namegt ltlocationgtSheraton waikiki
hotel Honolulu, hawaii, USAlt/locationgt
How about ltconfgtWWW2002 The eleventh
international world wide webconlt/confgt ltplacegtSher
aton waikiki hotel Honolulu, hawaii, USAlt/placegt
Then how about lt??gtWWW2002 The eleventh
international world wide webconlt/??gt lt??gtSheraton
waikiki hotel Honolulu, hawaii, USAlt/??gt
8What Is Needed?
- A resource should provide information about
itself - also called metadata (data about data)
- Metadata capture part of the meaning of data
- metadata should be in a machine processable
format - agents should be able to reason about
(meta)data - metadata vocabularies should be defined
9What Is Needed (Technically)?
- To make metadata machine processable, we need
- unambiguous names for resources (URIs)
- a common data model for expressing metadata (RDF)
- and ways to access the metadata on the Web
- common vocabularies (Ontologies)
- The Semantic Web is a metadata based
infrastructure for reasoning on the Web - It extends the current Web (and does not replace
it)
10Ontology Origins and History
- Ontology in Philosophy - a philosophical
disciplinea branch of philosophy that deals with
the nature and the organization of reality - Science of Being (Aristotle, Metaphysics, IV, 1)
- studies being or existence as well as the basic
categories thereof - trying to find out what entities and what types
of entities exist - has strong implications for the conceptions of
reality.
11Ontology in Computer Science
- An ontology is an engineering artifact
Neches91 - defines basic terms and relations comprising the
vocabulary of a topic area - the rules for combining terms and relations to
define extensions to the vocabulary - An explicit specification of a
conceptualization Gruber93 - Formal specification of a shared
conceptualization (of a certain domain) Borst
97 - Shared understanding of a domain of interest
- Formal and machine manipulable model of a domain
of interest
12Ontology Elements
- Concepts (classes) their hierarchy
- Concept properties (slots / attributes)
- Property restrictions (type, cardinality, domain,
etc.) - Relations between concepts (disjoint, equality,
etc.) - Instances
- E-R diagram / UML diagram ???
- Note Property ? Slot ? Relation ?
Relationtype ? Attribute ? Semantic link
type
13Ontology Languages
- RDF Schema
- RDF is a data model for objects and relations
between them - RDF Schema is a vocabulary description language
- Describes properties and classes of RDF resources
- Provides semantics for generalization hierarchies
of properties and classes
14Web Ontology Languages (2)
- OWL
- A richer ontology language
- relations between classes
- e.g., disjointness
- cardinality
- e.g. exactly one
- richer typing of properties
- characteristics of properties (e.g., symmetry)
- Logic
- BOTH are standards of www.w3.org
15History of the Semantic Web
- Web was invented by Tim Berners-Lee (amongst
others), a physicist working at CERN - TBLs original vision of the Web was much more
ambitious than the reality of the existing
(syntactic) Web - TBL (and others) have since been working towards
realising this vision, which has become known as
the Semantic Web - E.g., article in May 2001 issue of Scientific
American
... a goal of the Web was that, if the
interaction between person and hypertext could be
so intuitive that the machine-readable
information space gave an accurate representation
of the state of people's thoughts, interactions,
and work patterns, then machine analysis could
become a very powerful management tool, seeing
patterns in our work and facilitating our working
together through the typical problems which beset
the management of large organizations.
16Adding Semantics
- External agreement on meaning of annotations
- E.g., Dublin Core (http//dublincore.org/)
- Agree on the meaning of a set of annotation tags
- Problems with this approach
- Inflexible
- Limited number of things can be expressed
- Use Ontologies to specify meaning of annotations
- Ontologies provide a vocabulary of terms
- New terms can be formed by combining existing
ones - Meaning (semantics) of such terms is formally
specified - Can also specify relationships between terms in
multiple ontologies
17Berner-Lees Architecture
? Semanticsreasoning
?
? Relational Data
?
? Data Exchange
- Relationship between layers is not clear
- OWL DL extends DL subset of RDF
18The Role of Ontologies on the Web
- Ontologies provide a shared understanding of a
domain semantic interoperability - overcome differences in terminology
- mappings between ontologies
- Ontologies are useful for the organization and
navigation of Web sites - Ontologies are useful for improving the accuracy
of Web searches - search engines can look for pages that refer to a
precise concept in an ontology - Web searches can exploit generalization/
specialization information - If a query fails to find any relevant documents,
the search engine may suggest to the user a more
general query. - If too many answers are retrieved, the search
engine may suggest to the user some
specializations. - General e-business automation based on
understanding web resource in order to facilitate
intelligent (software agent) processing
19Case study Use of Ontology in an e-Marketplace
- D.K.W. Chiu, J.K.M. Poon, W.C. Lam, C.Y. Tse,
W.H.T. Siu, W.S. Poon. How Ontologies Can Help in
an E-marketplace, European Conference on
Information Systems 2005 (ECIS 2005), May 2005 - Semantic Web vision is probably too ambitious
- A more realistic current application that has a
potential to become a killer application
20Motivation
- Compare some general-purposed e-Marketplaces
(auction based) - e-Bay (HK) www.ebay.com.hk
- Yahoo Auction (HK) auctions.yahoo.com.hk
- Taobao owned by Alibaba.com http//www.taobao.com
- (See also Alibaba.com http//china.alibaba.com/)
- Compare special-purposed e-Marketplaces
- Airtickets http//www.qunar.com/
- Finding friends (!) http//hk.personals.yahoo.com
/ - Which one is better? Why?
- Key issue gt capturing and applying domain
knowledge
21What is an e-Marketplace?
e
-
Marketplace
offers
Aggregate requests
Repository
from Buyers, contact
bids
potential Suppliers,
Ontologies and Concepts
match Suppliers
e
-
Negotiation data
and Buyers, exchange
offers
Agreements
-
bids and offers,
generate e
-
Contract
bids
Buyers
22Problem Statements
- Are there currently significant practical use of
the Ontology from Semantic Web? - Match-making and beyond
- Software requirement engineering / negotiation
- Model and solve practical problems with CS ICT
- Cross-over multi-disciplinary research
- IJSSOE Dickson Chiu, Editor-in-chief
- http//www.igi-global.com/journals/details.asp?id
34268
23Example Ontology Clothing and Sales Negotiation
24Objective and Solution Approach
- How to elicit negotiation requirements?
- Semantic Web gt Ontologies
- gt help negotiators mutual understanding of
- issues, alternatives, and tradeoffs
- Address semantic requirements of negotiation
- Reduce cost and improve effectiveness of
negotiation(avoid combinatorial explosion of
issues) - Development of an effective and efficient
negotiation plan - Applications e-Marketplace, Web-service
negotiation, agent negotiation, requirement
negotiation
25Semantic basede-Marketplace Conceptual Model
26Overall e-Negotiation Process Design Methodology
Requirements elicitation phase
Decision phase
27Requirement Elicitation Methodology
- Traders select agreed ontology.
- Traders relate requirements to concepts in the
selected ontology. - System checks dependencies of concepts that
constitute all the requirements from the
(refined) ontology map. Mutually dependent
clusters of concepts determine the indivisible
groups of requirements that have to be considered
together so that effective tradeoff can be
evaluated. - The system checks the consistency of all the
concepts, issues, and their dependencies (Cheung
et al. 2002). - For a consistent plan, the system can proceed to
elicit the possible alternatives otherwise we
have to re-iterate from step 3. - According to the dependencies, the system can
formulate a precedence graph of the requirements
and requirements groups. Based on the precedence
graph, an efficient decision plan can be
determined.
28Decision Phase Methodology
- The system
- searches for the matching offers based on the
traders preference - attempt to rank them for the trader to choose
- Trader may accept any matched offers
- or change his reservation price and attempt a
negotiation with those offers in order to seek
for a more favorable one. - If no matching offers are found, the system
identifies near misses and also attempts to rank
them for the trader to choose. - Trader change his mind to accept a near miss
- or choose a near miss for negotiation.
- During negotiation, the system supports the user
to make and evaluate offers / counter-offers
based on the decision plan (from previous slide)
in a negotiation session as follows (Chiu et al.
2005). - Should new requirement issues arise in the
decision phase (say, due to incomplete
specification), the trader can we can go back to
analyze the new issue and its relationships to
the existing ones. - In real-life, the formulation of a decision plan
may involve several iterations. This reflects the
traders may not be able to understand all the
inter-relationships among the issues in one shot.
29Understanding Requirements from Ontologies
- Perform graph search algorithm on the semantic
map - Key requirements are preliminary identified in
the first round (e.g., unit price, quantity) - For each identified requirement issue,
- check if an issue can be mapped directly to a
concept. - If not, see if an issue can be refined into a set
of more specific concepts - a cost is refined into constituent costs that sum
up to it. - Incomplete Ontologies
- Introduce new concepts into the ontology map
- Relate it with to existing ones
30Understanding Requirements from Ontology (Cont)
- Perform graph search algorithm on the semantic
map - For each identified concept c,
- Examine every un-visited node n adjacent to c in
the ontology map. - For each such node n, see if the new concept is
relevant to the negotiation problem. - Repeat until no more related new concepts can be
identified. - Only after successful deal do we need to consider
combining newly identified working concepts back
to more concise real-life objects in specifying a
agreement - E.g., component costs need not shown to business
partner
31Understanding Dependencies of Requirements from
Ontologies
- Functional dependency
- borrowed from fundamental relational database
concepts - motivate this research
- The alternative for an issue is determined by the
alternatives(s) of other issue(s). - E.g., delivery date and quantity -gt cost of
production - Computational dependency
- more obvious type of functional dependency
- hardwired computational formula
- E.g., insurance amount percentage cost of
goods.
32Understanding Dependencies of Requirement from
Ontology
- Requirement dependency (constraint satisfaction)
- Only after the determinant value is known can
viable alternatives be determined. - E.g., whether a customer may pay by credit card,
bank draft, or remittance is evaluated according
to the total amount. - Classification dependency
- A special type of requirement dependency in which
the classification of another issue is dependent
on the outcome of an agreed issue. - E.g., customer tiering
33Indivisible Requirement Components for Tradeoff
Evaluation
- Indivisible Components of Issues
- Cyclic dependencies among the concepts
- Tradeoff Evaluation
- Topological sort of semantic graph gives
negotiation plan
34Understanding Possible Requirement Alternatives
from Ontology
- Alternative for requirements are often in
discrete values - cannot be expressed in numerical values
- not quantized in normal practices because of
difficulties in recognizing them, e.g., color - for simplicity and convenience (size gt S, M, L,
XL) - The elicitation of options is streamlined when a
complicated issue is decomposed into
concepts(appearance gt size color shapes) - Ontology provide
- explicit ordering of them (size gt S lt M lt L lt
XL) - implicit ordering
- inheritance (is-a) hierarchies
- composition hierarchies
35Exploring more trading opportunities from
Ontology
- Improve the accessibility of automated agents to
match functional specification - Intelligent software agents could represent
buyers or sellers - e-marketplace acts as broker
- Consider shared ontology attributes and
constraints - Map for cross-sale
- Group buyers or sellers together for higher
market efficiencies - Better hints for data mining
36 System Implementation Architecture
37OWL Listing
- ltrdfrest rdfresource"http//www.w3.org/1999/02/
22-rdf-syntax-nsnil"/gt - ltrdffirst rdfdatatype"http//www.w3.o
rg/2001/XMLSchemastring"gtSmalllt/rdffirstgtlt/rdfL
istgtlt/rdfrestgt - ltrdffirst rdfdatatype"http//www.w3.org/2001/X
MLSchemastring"gtMediumlt/rdffirstgtlt/rdfListgtlt/rd
frestgt - ltrdffirst rdfdatatype"http//www.w3.o
rg/2001/XMLSchemastring"gtLargelt/rdffirstgtlt/rdfL
istgtlt/rdfrestgt - ltrdffirst rdfdatatype"http//www.w3.org/
2001/XMLSchemastring"gtExtra Largelt/rdffirstgtlt/rd
fListgt - lt/owloneOfgtlt/owlDataRangegtlt/rdfsrangegt
- lt/owlDatatypePropertygt
- ltowlClass rdfID" UnitCost"gt
- ltowlequivalentClassgt lt!-- unit cost depends
on appearance --gt - ltowlRestrictiongt ltowlsomeValuesFrom
rdfresource"Appearance" /gt lt/owlRestrictiongt - lt/owlequivalentClassgt
- lt/owlClassgt
- lt/owlOntologygt
- ltowlOntology rdfabout"Clothing"gt
- ltrdfscommentgtSample Clothing
Ontologylt/rdfscommentgt - ltowlClass rdfID"Clothing" /gt
- ltowlClass rdfID"Appearance" /gt
- ltowlClass rdfID"Color"gt
- ltrdfssubClassOf
rdfresource"Appearance" /gt - ...
- lt/owlClassgt
- ltowlObjectProperty rdfID"hasAppearance"
gt - ltrdfsdomain rdfresource"Clothi
ng" /gt - ltrdfsrange rdfresource"Appeara
nce" /gt - lt/owlObjectPropertygt
- ltowlObjectProperty rdfID"hasColor"gt
- ltrdfssubPropertyOf
rdfresource"hasClothAppearance" /gt - ltrdfsrange rdfresource"Color
/gt - ...
- lt/owlObjectPropertygt
- ltowlDatatypeProperty rdfID"size"gt lt!--
Enumeration --!gt - ltrdfsdomain rdfresource"Appearance"/gt
38Summary
Function Traditional e-marketplace problem Contributions of Ontology
Match-making Match-making is often ineffective because of the rigid definition of products of limited attributes. Shared and agreed ontology provides common, flexible, and extensible definitions of products and requirements for match-making and subsequent business processes
Match-making It is difficult to specify complex product requirements because the relationships among attributes and values are ignored. Complicated requirements can be decomposed into simple concepts for streamlining the elicitation of options
Match-making User interactions are limited to mainly manually, which is time consuming. Accessible by automated agents through Semantic Web specifications for more business opportunities
Recom-mendation Recommendations are often only possible within the same category. Ontology helps elicit alternatives for recommendation.
Recom-mendation Pre-set formulae for every type of product are needed for evaluation. Ontology help recommendation by evaluating offers in terms of flexible overall scaling
Recom-mendation Cross-sale and grouping of buyers and sellers with similar requests are difficult. Matching grouping of buyers and sellers as well as cross-sale possible by inference with the ontology.
Negotiation No implicit ordering of alternatives. Implicit ordering of alternatives is elicited via inheritance.
Negotiation Manual negotiation or inadequate negotiation support cause inefficient process and ineffective recognition. Machine understandable semantics facilitate negotiation and automatic configuration of products and services as specified.
39Conclusions
- Formulation of negotiation plan with maturing of
Semantic Web technologies - Elicitation of negotiation issues, issue
dependencies, tradeoff, and alternatives - Control the openness of issues
- Our algorithm verifies the completeness of
elicited negotiation requirements - Negotiation processes are properly guided,
recorded, and managed - For e-commerce activities are usually more
structural and repeatable (as opposed to
political negotiations) - Ontologies and plans are therefore reusable
- Negotiation automation with agents / integration
with EIS
40Future Work
- Formal models
- Elicitation of semantic distances
- enhancement of ontology-based matchmaking and
recommendation algorithms - ontology-based cross-sale and up-sale
- grouping of buyers and sellers for combined
quantity deals - mobile clients and constraint-based requirement
specification
41Question and Answer
Thank you! Email dicksonchiu_at_ieee.org