Netcentric Serviceoriented Enterprises

1
Net-centric Service-oriented Enterprises

• Bina Ramamurthy
• Chapter 1 of The Semantic Web book

2
Introduction

• Werservices? Mashups ?Service-oriented
  architectures? What next?
• Web is still a set of static and dynamically
  generated web pages linked together.
• Usually coded in html and meant for human
  consumption.
• Web information need to be used not only display
  purposes but also for interoperability and
  integration between systems and applications XML
  is a solution that partially addresses this need.
• In order to enable machine-machine exchange and
  automated processing we need to provide
  information in such a way that machines can
  understand.
• New standards and languages are being
  investigated and developed to give meaning to web
  information.
• Examples RDF (Resource Description framework),
  OWL (Web Ontology Language)
• Improve expressiveness of the web, allow
  automatic and semiautomatic processing of web
  resources and web pages.
• Answer to What next? Semantic Web Services for
  a given business/industrial domain

3
Topics for Discussion

• Creating and using semantic information
• Ontologies Cornerstone of Semantic Web Services
  and service-oriented enterprises
• Characteristics of a new world
• Challenges for service-based applications
• Importance of semantics for organizations
• Semantic Service Oriented Architecture
• Ontologies Ontology management system

4
Semantic information

• Discover, acquire, and create metadata for
  unstructured, semi-structured, and structured
  information
• Reason, interpret, infer and answer using
  semantics
• Represent, organize, integrate resources, content
  and knowledge using semantics
• provision, present, communicate and act using
  semantics
• Provide machine-machine semantic interface,
  human-semantics semantic interface

5
WS and SOA and Semantics (Web)
HTML
W
XML
WS
SOA
SWS
SW
OWL
6
Ontologies in Business

• Provide formal support for communication between
  agents and exchange of knowledge.
• In the context of human communications, it aims
  at reducing and eliminating terminological and
  conceptual confusion.
• Unifying framework enabling cooperation amongst
  people in reaching better inter-enterprise
  organization.

7
Examples of successful Ontologies

• Healthcare hierarchical and controlled vocabulary
  for human disease representation
• Food and agriculture organizations of the United
  Nations FAO
• Data management and interchange between
  enterprises Open EDI for business transactions
• Scientific Computing
• Knowledge representation ontologies.

8
Characteristics of a new world

• Information and knowledge are key enablers of
  business and economic performance and sustainable
  development.
• Globalization creation and consumption of
  knowledge and information are made in the global
  context.
• Exploitation of synergies and capacities beyond
  boundaries
• Realization of new opportunities
• Understanding of threats
• Human and social networks
• New levels of performance

9
Characteristics Business networking

• Business and economic activities as well as
  competition, require new models of business
  networking.
• Advanced documentation of skills and competencies
• Newer business models Example IF.coms banking
  product
• Context-based collaboration define new demands
  for advanced business networking at global level

10
Characteristics Shared Models

• A global consensus toward peace, development,
  health and prosperity needs to be based on shared
  conceptual models that addresses issues on global
  scale. Examples
• Global warming
• Mars exploration
• Financial and manufacturing sectors
• Global AIDS initiatives
• Global information landscape shared models are
  required for interoperability, exploitation of
  collective intelligence.

11
Characteristics Collective intelligence

• Apply collective intelligent filters or
  collaborative filtering in the context of global
  information world.
• These may challenge the traditional models of
  business performance, marketing and
  profitability.
• Example Financial domain once again revisit
  IF.com

12
Characteristics Open Paradigm

• Open paradigm relates with several complimentary
  movements
• Open source software
• Open content
• Open access
• Open knowledge
• Open research
• Open culture
• Result amazing capacity to support new business
  models and several application models.
• Example amazon EC2, cloud computing, map reduce

13
Challenges for Semantic Web Services-based
Systems Summary

• Definition of new modes of human, knowledge and
  business networking beyond local boundaries well
  defined conceptual models that match information
  sources and human services.
• With ontologies and social networks as anchors
• Process and service-oriented infrastructure
• Globalizing information and definition of new
  contexts for value exploitation
• Design of multiple reference levels to the same
  set of information and knowledge delivers a new
  level for dynamic, and personalized systems.
• Internationalization

14
Challenges

• Delivering and integrating quality to
  information
• Enormous explosion of content while quality is
  very subjective concept
• We need infrastructure that deliver assessment
  models of information quality.
• Integration of isolated information assets build
  more meaningful services. SOA can help in this
  aspect. Lets discuss how?
• Support of business value and co-located
  distributed business models crucial aspect is to
  translate web semantic ontologies to business
  models. SOA can help bridge this gap. How?
• Promotion of a critical shift in human
  understanding and interacting with digital world
  Web needs to respond to human demand for richer
  modes of meaningful, useful and productive
  interaction. Combination of semantics and SOA can
  help here.

15
Importance of semantics and services (SOA) to
organizations

• Integration is the top priority for many
  worldwide enterprises.
• Inter, intra and human interface integration.
• Cross-organization cooperation in small and
  medium enterprises (SME).
• Semantics and service combination can facilitate
  discovery of heterogeneous components, data
  integration and communication.
• Semantics SOA is most suitable for
  business-business interaction and in integration
  of e-business value chains. Ex amazon.com market
  place, yahoo.com

16
Ontology Management System (OMS)

• Supports entire lifecycle of inter-enterprise
  ontologies, including creation, storage, search,
  query, reuse, maintenance, and integration.
• An OMS needs to address a wide range of problems
  ontology models, ontology base design, query
  languages, programming interfaces, query
  processes and optimization, federation of
  knowledge sources, caching and indexing,
  transaction support, distributed system support,
  and security support.

17
Overall Challenge

• High volume and wealth of data and information
  generated by the numerous web applications that
  needs to be analyzed and processed to provide
  useful and timely knowledge for decision makers.
• Arcelor Mittal 330000employees, 60 countries,
  flat steel products. How to extract knowledge
  from the information generated?

18
Contributions of the text chapters

• Important industries (vertical domains) covered
  semantic enterprises, finance, government,
  healthcare and life sciences, education, business
  and customer management, enterprise management
  and security.
• Highlights in the context of actual industry, the
  full range of business and technological issues
  that must be addressed.
• Provides a comprehensive discussion of the
  required integration of semantic web services
  (SOA) and business strategies.
• Sets a context for critical thinking.

19
Project ideas

• Read the text
• Choose a vertical domain that appeals to you and
  that is familiar to you.
• Form your project group of two.
• More directions will be given in Assignment1

20
Chapter 2 Semantic Enterprises

• For a concept to be widely adopted it needs to
  reach a level of maturity.
• Semantic web is a new concept that still has some
  distance to go before it reaches a point of this
  widespread adoption.
• Lets examine how semantic web (tools and
  technologies) can help address some of the
  challenges that companies are facing today.

21
Topics for Discussion

• The Business context
• Tools and Technologies for representing semantics
• Software for semantic services
• Use cases for semantic representation of
  information
• Recruitment services
• Agile manufacturing
• Patterns and insights in data
• Integration of scientific data
• Enterprise search and navigation
• Compliance and regulation
• SOA metadata

22
Business and Technology Drivers The Context

• Read Section 2 of Ch.2
• Commercial organizations are always under
  pressure to perform financially.
• Growing interest in being able to integrate all
  data related to the core components that drive
  their success.
• Integrate not only structured data but also huge
  volume of unstructured data collected and
  generated. Ex explosive email
• Many industries are moving towards collaborative
  business models. Ex drug discovery and clinical
  trials
• Companies conduct businesses in many countries.
• Integrating data across department also comes
  with its challenges.
• Ability to respond rapidly to change.
• Data is the most important asset and access to it
  should be controlled. Provide API for access and
  build revenue models around it.
• Effective use of business data and change/adapt
  as needed.

23
Tools and Technologies RDF

• Resource Description Framework (RDF) is a core
  semantic web recommendation from W3C.
• Represents data using triplet subject-predicate-o
  bject

24
Tools and Technologies OWL

• OWL (Web Ontology Language) is a more expressive
  language once a a standard from W3C.
• It provides ways to define classes and instances
  and relationship for modeling real-world objects.
• ltowlClass rdfID"PotableLiquid"gt
  ltrdfssubClassOf rdfresource"ConsumableThing"
  /gt ... lt/owlClassgt
• SPARQL is a query language for RDF and OWL.

25
Software

• Databases, middleware and applications must be
  enhanced to work with RDF, OWL and SPARQL.
• Remember most of todays data is in relational
  databases and in XML formats. So we need
  converters or interfaces to bridge this gap.
• Pages 21-22 has a excellent collection of
  software initiatives in this direction.
• Bottom line is that we need to pay attention to
  data representation in order to build an
  efficient SOA.
• On to use cases..

26
Recruitment Services
hireme.com
Geo4930956
hiremecandidate101
hiremecandidate102Job1
hiremecandidate102Job2
hiremecandidate102Job3
RDF..
27
Recruitment Services

• To enable querying across multiple recruiter
  databases, the hiring company would encourage all
  its recruitment agencies to make a subset of
  their data available in RDF.
• Common vocabulary, SPARQL endpoint
• OWL could also be used
• Conversion to legacy relational info into OWL or
  RDF
• Expose recruitment companies services too.

28
Agile Manufacturing

• Business drivers
• Reuse machines
• Use ingredients in multiple products
• Follow trends and latest food craze
• FDA labeling regulations
• Enterprise resource planning incorporate
  ontologies
• Scheduling supported by food ontologies.
• Services and semantics will make it easier to
  incorporate new data that is deemed relevant and
  help in decision making.

29
Identifying of Patterns and Insights in data

• Business drivers
• Non-structured data reports, email
• Need to mine this data
• Use natural language to extract triplets and
  store as RDF which can then be queried.
• Association of semantics and services will make
  the querying this RDF or OWL database very
  efficient.
• Oracle database supports RDF and OWL.
• Java APIs are available for querying patterns.

30
Integration of Scientific Data

• Business driver
• Drug discovery and development is very expensive
  and time consuming process
• For a drug to get from bench to market takes 5000
  screened compounds, 15 years and nearly 1
  billion.
• Desire to eliminate late stage attrition
  identifying and eliminating drugs that do not
  have the desirable safety profiles.
• Need to be aware of competitive offerings or
  patents to access market potential
• Solution semantic data integration of
  heterogeneous databases and services for semantic
  queries

31
Integration of scientific data (contd.)

• Data types include chemical structures,
  biological sequences, images, biological
  pathways, clinical observations and scientific
  papers.
• Data warehouse is NOT a solution.
• We need a unified view with no ambiguity in
  terms.. GSK protein needs to different than GSK
  the company name..
• This would allow biological mashups for discovery
  and decision making.

32
Enterprise Search and Navigation

• OTN (The Oracle Technology Network) is the main
  source of technical information for oracle
  developer community.
• Web site provides access to product
  documentation, product releases, software
  downloads, etc. Richness, complexity and dynamism
  of the information made it challenging for
  traditional search.
• Oracle worked with Siderean and created a
  semantic web http//otnsemanticweb.oracle.com

33
Compliance and Regulation

• Increasing complex set of regulations by such
  congressional acts such as SARBOX and HIPPA.
• Policies can be implemented using semantic web.
• Semantics and services can keep trace and verify
  compliance.

34
SOA Metadata

• Semantics can be used to assign metadata that
  will help in true dynamic discovery, invocation
  and composition.
• Thus semantics can improve inherent flexibility
  of SOA infrastructure.

35
Summary

• We understand that incorporating semantics into
  the services infrastructure can help advance SOA
  goals.
• Future designs should consider both semantic web
  concepts and SOA concepts.