Explaining The Semantic Web


Published on

Complete details explaining Semantic Web

Published in: Technology

Explaining The Semantic Web

  1. 1. Making Sense of the Semantic Web Nova Spivack CEO & Founder Radar Networks Radar Networks
  2. 2. About This Talk • Making sense of the semantic sector • How the Semantic Web works • Future outlook • Twine.com • Q&A Radar Networks
  3. 3. The Big Opportunity… The social graph just connects people The semantic graph connects everything People Companies Emails And it uses richer semantics to enable: Places Products Better search Interests Services More targeted ads Smarter collaboration Activities Web Pages Deeper integration Projects Documents Richer content Events Multimedia Better personalization Groups Radar Networks
  4. 4. The third decade of the Web • A period in time, not a technology… • Enrich the structure of the Web o Improve the quality of search, collaboration, publishing, advertising o Enables applications to become more integrated and intelligent • Transform Web from fileserver to database o Semantic technologies will play a key role Radar Networks
  5. 5. The Intelligence is in the Connections Intelligent Web Web OS Web 4.0 2020 - 2030 Intelligent personal agents Semantic Web SWRL Web 3.0 Distributed Search OWL 2010 - 2020 OpenID AJAX SPARQL Semantic Databases between Information Social Web RSS ATOM Semantic Search Widgets P2P RDF Mashups Office 2.0 Javascript SOAP XML Flash Web 2.0 The Web Java 2000 - 2010 Weblogs Social Media Sharing HTML HTTP SaaS Social Networking Directory Portals Wikis VR Keyword Search Lightweight Collaboration The PC BBS Gopher Web 1.0 Websites 1990 - 2000 MMO’s MacOS SQL Groupware SGML Databases Windows File Servers The Internet PC Era FTP IRC Email 1980 - 1990 USENET PC’s File Systems Connections between people Radar Networks
  6. 6. Beyond the Limits of Keyword Search The Intelligent Web Web 4.0 ctivity of Search 2020 - 2030 Reasoning The Semantic Web Web 3.0 Semantic Search 2010 - 2020 The Social Web Natural language search The World Wide Web Web2010 2000 - 2.0 Tagging Web2000 1990 - 1.0 Keyword search The Desktop Directories PC Era 1980 - 1990 Files & Folders Databases Amount of data Radar Networks
  7. 7. Five Approaches to Semantics • Tagging • Statistics • Linguistics • Semantic Web • Artificial Intelligence Radar Networks
  8. 8. The Tagging Approach • Pros • Technorati o Easy for users to add and read tags • Del.icio.us o Tags are just strings o No algorithms or ontologies to • Flickr deal with o No technology to learn • Wikipedia • Cons o Easy for users to add and read tags o Tags are just strings o No algorithms or ontologies to deal with o No technology to learn Radar Networks
  9. 9. The Statistical Approach • Pros: • Google o Pure mathematical algorithms o Massively scaleable • Lucene o Language independent • Autonomy • Cons: o No understanding of the content o Hard to craft good queries o Best for finding really popular things – not good at finding needles in haystacks o Not good for structured data Radar Networks
  10. 10. The Linguistic Approach • Pros: • Powerset o True language understanding o Extract knowledge from text • Hakia o Best for search for particular facts or relationships • Inxight, Attensity, and others… o More precise queries • Cons: o Computationally intensive o Difficult to scale o Lots of errors o Language-dependent Radar Networks
  11. 11. The Semantic Web Approach • Pros: • Radar Networks o More precise queries o Smarter apps with less work • DBpedia Project o Not as computationally intensive o Share & link data between apps • Metaweb o Works for both unstructured and structured data • Cons: o Lack of tools o Difficult to scale o Who makes all the metadata? Radar Networks
  12. 12. The Artificial Intelligence Approach • Pros: • Cycorp o This is the holy grail!!!! o Approximates the expertise and common sense reasoning ability of a human domain expert o Reasoning / inferencing, discovery, automated assistance, learning and self-modification, question answering, etc. • Cons: o This is the holy grail!!!! o Computationally intensive o Hard to program and design o Takes a long time and a lot of work to reach critical mass of knowledge Radar Networks
  13. 13. The Approaches Compared Make the Data Smarter A.I. Semantic Web Linguistics Tagging Statistics Make the software smarter Radar Networks
  14. 14. Two Paths to Adding Semantics • “Bottom-Up” (Classic) o Add semantic metadata to pages and databases all over the Web o Every Website becomes semantic o Everyone has to learn RDF/OWL • “Top-Down” (Contemporary) o Automatically generate semantic metadata for vertical domains o Create services that provide this as an overlay to non-semantic Web o Nobody has to learn RDF/OWL -- Alex Iskold Radar Networks
  15. 15. In Practice: Hybrid Approach Works Best Tagging Semantic Web Top-down Statistics Linguistics Bottom-up Artificial intelligence Radar Networks
  16. 16. A Higher Resolution Web IBM.com Web Site Joe Person Lives in Palo Alto IBM City Company Publisher of Fan of Subscriber to Lives in Employee of Sue Jane Person Dave.com Person RSS Feed Fan of Coldplay Band Friend of Member of Depiction of Design Married to Source of Team Member Group 123.JPG of Dave.com Bob Photo Weblog Person Depiction of Member of Dave Stanford Member of Author of Person Alumnae Group Member of Radar Networks
  17. 17. The Web IS the Database! Application A Application B IBM.com Web Site Joe Person IBM Palo Alto Lives in City Company Publisher of Fan of Subscriber to Lives in Employee of Sue Jane Person Dave.com Person RSS Feed Coldplay Fan of Band Friend of Member of Design Depiction of Team Married to Group Source of Member 123.JPG of Photo Dave.com Bob Weblog Person Depiction of Member of Dave Stanford Person Alumnae Member of Author of Group Member of Radar Networks
  18. 18. Smart Data • Smart Data is data that carries whatever is needed to make use of it: • Software can become dumber and more generic, yet ultimately be smarter • The smarts moves into the data itself rather than being hard-coded into the software Radar Networks
  19. 19. The Semantic Web is a Key Enabler • Moves the “intelligence” out of applications, into the data • Data becomes self-describing; Meaning of data becomes part of the data • Data = Metadata. • Just-in-time data • Applications can pull the schema for data only when the data is actually needed, rather than having to anticipate it Radar Networks
  20. 20. The Semantic Web = Open database layer for the Web User Web Ads & Data Apps & Profiles Content Listings Records Services Open Query Interfaces Open Data Mappings Open Data Records Open Rules Open Ontologies Radar Networks
  21. 21. Semantic Web Open Standards • RDF – Store data as “triples” • OWL – Define systems of concepts called “ontologies” • Sparql – Query data in RDF • SWRL – Define rules • GRDDL – Transform data to RDF Radar Networks
  22. 22. RDF “Triples” Predicate Subject Object • the subject, which is an RDF URI reference or a blank node • the predicate, which is an RDF URI reference • the object, which is an RDF URI reference, a literal or a blank node Source: http://www.w3.org/TR/rdf-concepts/#section-triples Radar Networks
  23. 23. Semantic Web Data is Self-Describing Linked Data Ontologies Definition Definition Definition Definition Data Record ID Definition Field 1 Value Field 2 Value Definition Field 3 Value Field 4 Value Definition Radar Networks
  24. 24. RDBMS vs Triplestore Person Table S PO Subject Predicate Object ID f_name l_name 001 isA Person 001 firstName Jim 001 jim wissner 001 lastName Wissner 002 nova spivack 001 hasColleague 002 003 chris jones 002 isA Person 002 firstName Nova 004 lew tucker 002 lastName Spivack 002 hasColleague 003 003 isA Person 003 firstName Chris Colleagues Table 003 lastName Jones 003 hasColleague 004 SRC-ID TGT-ID 004 isA Person 001 001 004 firstName Lew 001 002 004 lastName Tucker 001 003 001 004 002 001 002 002 002 003 002 004 003 001 003 002 003 003 003 004 004 001 004 002 004 003 004 004 Radar Networks
  25. 25. Merging Databases in RDF is Easy S PO S P O S PO Radar Networks
  26. 26. The Growing Linked Data Universe Twine Yahoo Freebase Reuters OpenCalais Radar Networks
  27. 27. The Growing Semantic Web Online Services Consumers Developers Applications Radar Networks
  28. 28. Future Outlook • 2007 – 2009 o Early-Adoption o A few killer apps emerge o Other apps start to integrate • 2010 – 2020 o Mainstream Adoption o Semantics widely used in Web content and apps • 2020 + o Next big cycle: Reasoning and A.I. o The Intelligent Web o The Web learns and thinks collectively Radar Networks
  29. 29. The Future of the Platform… • 1980’s -- The Desktop is the platform • 1990’s -- The Browser / Server is the platform • 2000’s -- Web Services are the platform • 2010’s -- The Semantic Web is the platform • 2020’s -- The WebOS is the platform • 2030’s -- The Human Body is the platform…? Radar Networks
  30. 30. A Mainstream Application of the Semantic Web… Radar Networks
  31. 31. Twine.com Overview Organize. Share. Discover. Around your interests Using the Semantic Web Radar Networks
  32. 32. What Can You Do With Twine? • Organize o Collect & manage your stuff • Share o Author & share content o Discuss & collaborate • Discover o Track Interests o Search & explore o Get recommendations Radar Networks
  33. 33. Differentiation • Facebook - For your relationships • LinkedIn - For your career • Twine - For your interests Twitter + Del.icio.us + Blogger? Radar Networks
  34. 34. Twine is Smart Semantic tagging Semantic linking Organize All Kinds Of Share Content Discover Recommendations Semantic Search Radar Networks
  35. 35. Let’s take a look at Twine… (demo of Twine site…) Radar Networks
  36. 36. Radar Networks’ Semantic Web Platform Web App Twine.com User Portal REST API Bookmarklet RSS Feeds Cache SPARQL & Email AJAX, Jetty, PicoContainer, Java, XML, SPARQL Jena, ATOM KnowledgeBase Semantic Object Class inferencing Object Query Tuple Cache & Cache Query Platform RDF, OWLOntology TupleStore service SQL Query Access Control Predicate Remote Cache Generator Inferencing Access RDF, OWL, SQL Mina SQL Database WebDAV File Store Storage Relational database Flat File Store Postgres, webDAV, Isilon Solaris cluster Radar Networks
  37. 37. Target Customer Twine is for active users of the Web, including consumers and professionals, who create, find and share information about their interests Demographics: Interests: • 18 – 45 years old • Professional associations • Have many personal interests and hobbies • Alumni groups • Social connections are important – family, friends, colleagues • Social networks (Facebook, Plaxo, LinkedIn) • Americans with a household income of $100,000 or more • Volunteer organizations o Nearly 26 million such consumers used the Internet in • Groups based on interests (hobbies, health, sports, August 2003, spending an average of 27.6 hours online entertainment, culture, family, technology, user groups, etc.) -- more than any other income segment. • Participating/working in teams at organizations of all sizes o Consume an average of nearly 3,000 pages a month, almost 300 pages more than the average Internet user Radar Networks
  38. 38. Market Opportunities for Twine Individuals Groups, Teams and Communities • Individual consumers • Interest communities • Support groups • Individual professionals • Content publishers • Users groups • Hobbyists • Social groups • Product communities • Event communities • Communities of practice • Customer support • Collaborative teams Radar Networks
  39. 39. Contact Info • Visit www.twine.com to sign up for the invite beta wait-list • You can email me at nova@radarnetworks.com • My blog is at http://www.mindingtheplanet.net • Thanks! Radar Networks
  40. 40. Rights • This presentation is licensed under the Creative Commons Attribution License. o Details: This work is licensed under the Creative Commons Attribution 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by/3.0/ or send a letter to Creative Commons, 171 Second Street, Suite 300, San Francisco, California, 94105, USA. • If you reproduce or redistribute in whole or in part, please give attribution to Nova Spivack, with a link to http://www.mindingtheplanet.net Radar Networks