Is The Semantic Web Ready for Business

1
Is The Semantic Web Ready for Business?
Information Systems Engineering Area Faculty of
Industrial Engineering and Management Technion
Israel Institute of Technology

• Eran Toch
• May 2005

2
Agenda

• Why do we need it?
• What is it?
• Challenges and Opportunities
• Applications
• Current Commercial Activities
• Semantic Web Services
• Social Software
• Summary

3
The WWW Technically

• A set of protocols and languages driven by a
  strong standards approach
• URI
• HTTP
• HTML
• XML
• Principles
• Implementation and platform independence crucial
• World Wide Web Consortium the most prominent

4
Problems with the WWW

• 65,900,000 results were returned
• Google - Market Cap72.45 B

5
More Problems Comparison Shopping

• Shopping.com - Market Cap502.70 M

6
Building Shopping Comparison Engine
7
Trust
Spam
Phishing
E-Commerce
Who can you trust to send you emails?
Is this site is the one it claims to be?
How can I know for sure if a transaction really
occurred?
8
Problem Domains

• The General Web
• Data-mining activities (e.g. search, comparison,
  notification)
• Transactions (e-comm, e-gov)
• Business Knowledge bases
• Intranets, data warehouses
• Collaborative Computing
• Transaction between systems
• Knowledge-based businesses
• biology, law etc

9
Agenda

• Why do we need it?
• What is it?
• Challenges and Opportunities
• Applications
• Current Commercial Activities
• Business Models
• Social Software
• Semantic Web Services
• Summary

10
The Semantic Web
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. Sir Tim Berners-Lee
SciAme
11
Making the Web Machine-Friendly

• Making knowledge self explainable for machines
• Creating an environment for knowledge inference
• Establishing trust

12
Making it Meaningful for Machines
ltRDFgt
WWW Resource
http//www.amazon.com/4344533
Humans
Machines
13
Knowledge For Machines
is a
RDF ResourceDescription Framework
Audio CD
Soundrack
itsType
Product
http//www.amazon.com/4344533
itsLabel
hasPrice
availability
Universal
14.99
hasTitle
24 hours
Kissing Jessica Stein
REF
14
Ontologies
Healthcare

• An ontology is standard for some knowledge
  domain, e.g.
• Healthcare
• Bioinformatics
• CRM
• Web services
• It provides a formal and agreed upon controlled
  vocabulary, which is used to define concepts
• Information can be tagged according to these
  concepts

Disease
Is treated by
Is a
Product
has
Medicine
Price
prescribe
takes
Doctor
Patient
treats
15
Web Ontology Language (OWL)

• OWL is an RDF-based language for Ontology
  modeling.
• Enable class and instance definition, using
  relations and properties such as
• Properties (price is a property of product)
• subClassOf (Employee is subClassOf Person)
• intersectionOf (music CD is intersectionOf
  playable thing and consumer product)
• Cardinality constrains (product has 1 (and only
  1) price properties)
• OWL ontologies can be developed independently,
  having concepts reference each other

16
Ontologies
E-Commerce
Healthcare
supplies
Supplier
Disease
has
Is treated by
Is a
Product
treats
RFID
Medicine
Doctor
has
Price
takes
buys
Patient
Customer
Is a
17
Generalizing Knowledge Networks
Product
Price Comparison Robot
Is a
Price
Medicine
Side Effect
Is a
Is a
Hospital Drug Monitor Robot
OTC Medicine
Prescription
18
The Network Effect
E-Commerce Site
Library
Is a
E-Commerce Site
Item
E-Commerce Site
Product
Price
Catalog ID
Is a
Personal Computer
Song
Path
19
Agenda

• Why do we need it?
• What is it?
• Challenges and Opportunities
• Applications
• Current Commercial Activities
• Semantic Web Services
• Social Software
• Summary

20
Business on the Semantic Web
Business Processes
21
Bottlenecks

• Sufficient metadata is the main bottleneck of the
  Semantic Web
• There is a loop
• Without metadata, no applications will be built
• Without applications, no one will create metadata

Commercial
The Metadata gap
Academic
22
Metadata Chasm

• Ontology creation requires companies and
  organization to standardize their concepts, much
  harder than to standardize communication
  protocols
• Ontology creation requires large investments.
  Because ontologies reduce the uncertainty of
  information, their benefits will be revealed
  mainly in the long run.
• Thus, they do not provide immediate return on the
  investment, not immediately KIM.
• However, in some markets, ontologies may have
  faster cost-to-benefit cycle.

23
Markets

• Niece fields on the general WWW
• Content syndication
• Communications and social networks
• Business Processes
• Handling interoperability
• Extending Service Oriented Architecture (SOA)
• Knowledge-rich markets
• Bioinformatics
• Software Engineering
• Law
• Business Intelligence
• Business intelligence using semantic annotations

24
Agenda

• Why do we need it?
• What is it?
• Challenges and Opportunities
• Applications
• Current Commercial Activities
• Business Models
• Social Software
• Semantic Web Services
• Summary

25
Semantic Web Applications

• Adobe - uses RDF as a basis for documenting
  meta-data, in PDF and other tools
• Boeing uses RDF and OWL in several internal
  projects
• AGFA uses RDF to categorize medical photos
• NOKIA lots of Semantic Web activities.
  Including RDF knowledge store
• IBM Strong research activities

26
Aduna

• Netherlands-based startup
• Main products
• Semantic desktop search
• Semantic enterprise search
• Semantic Metadata server

ADUNA
27
Celcorp

• Based in Santa-Monica
• Claim to have 3 customers from Fortune 500
• Main business Semantic EAI (Enterprise
  Application Integration)
• Main product Celware Intelligent Access
• Records user actions in legacy systems
• Builds an editable knowledge based of reusable
  task models
• Generates executable processes, based on the
  tasks
• Business processes are automatically planned and
  executed, using the knowledge base

CELCORP
28
Brandsoft

• Based in Los Gatos, CA
• Main product Brandsoft Resource Manager
• Content management and application development
  suite, based on RDF

29
Cerebra

• Based in Carlsbad, CA. AKA Network Inference.
• Claim to have more than 25 customers, some of
  them from Fortune 500
• Cerebra Server
• Provides ontology management and storage
• Semantic integration of data from RDBMS etc
• Querying through XQuery
• Cerebra Construct
• Ontology modeling using MS-Visio
• Professional Services

CEREBRA
30
Tucana

• Based in Reston, VA.
• Their main product, Tucana Knowledge Discovery
  Suite, is a semantic knowledge base, with some
  business intelligence abilities. It is built
  upon
• Scaleable RDF triple store
• Reasoning engine
• Metadata extraction from RDBMS, files, emails,
  ERP etc.

Tucana
31
Semantic Web Services
2. Domain and Service models can be used to
automatically or manually compose services
1. The customers agent automatically locates and
invokes the brokerage firms Web service
32
Semantic Web Services, contd

• OWL-S is an upper ontology for a semantic
  description of Web services.
• E.g. an input message can be typed as the concept
  Product, and not just a String
• Describes a Web service by
• What it does (inputs, outputs, preconditions)
• How it works (a process model)
• Grounding to an invocation method (WSDL)

OWL-S
33
Semantic SOA (Service Oriented Computing)
Employee Signoff
Orderproduct
Report employeedaily activities
Order Management
Update inventory
check order status
Check usersecurity profile
Customer Care
Get customerHistory
Get incomingmessages
34
FOAF

• Stands for "Friend Of A Friend
• Provides a template for metadata about people,
  and their interests, relationships and activities
• An open community-lead and open-source initiative

FOAF
35
FOAF Example
Email
Picture
ltPersongt
knows
ltPersongt
Website
knows
Name
ltPersongt
36
FOAF-Based Applications

• FOAF Explorer
• More
• Job search
• Dating
• Identity
• Security

http//xml.mfd-consult.dk/foaf/explorer/
37
FOAF-Applications contd

• FOAFNaut - lthttp//www.foafnaut.org/gt

38
Email Trust with FOAF
TRUST
39
Inferring Trust
trust
Email is blocked
Email is accepted
me
trust
trust
email
email
trust
trust
trust
Donttrust
40
Agenda

• Why do we need it?
• What is it?
• Challenges and Opportunities
• Applications
• Current Commercial Activities
• Business Models
• Social Software
• Semantic Web Services
• Summary

41
Summary
Widespread Use
Academiadesert
Where would the semantic web go
42
Key Points for Success

• Crossing the metadata chasm
• Automatic extraction of metadata in predefined
  domains
• Reducing the turn-on-investment cycle. Making
  ontologies useful, now
• Niece markets
• Bioinformatics
• Software engineering
• Business Processes
• Leveraging semantic markup with Web services and
  enterprise computing

43
Long Term Implications of a Success

• New professions
• Ontology editors
• Taggers
• Agents
• Many automated tasks (shopping, travel, dating)
• Bigger threats on human agents (travel agents,
  insurance agents)
• Business Processes
• IT missions change - from constructing
  applications to providing frameworks
• Work of operational personnel change from
  requirement definitions to business process
  modeling

44
References

• SciAme Berners-Lee, T., Hendler, J., Lassila,
  O., The Semantic Web, Scientific American,
  284(5), 2001, pp. 34-43.
• RDF http//www.w3.org/RDF, http//www.ilrt.bris.
  ac.uk/discovery/rdf
• OWL http//www.w3.org/TR/owl-guide
• KIM Kim, Henry M. (2002). "Predicting How
  Ontologies for the Semantic Web Will Evolve",
  Communications of the ACM, Vol. 45, No. 2, pp.
  48-54.
• FOAF http//www.foaf-project.org/
• ADUNA http//aduna.biz
• CELCORP http//www.celcorp.com
• CEREBRA http//cerebra.com
• OWL-S http//www.daml.org/services
• TRUST http//trust.mindswap.org

45
Thank You
erant_at_tx.technion.ac.il http//www.technion.ac.il
/erant