Nova Spivack CEO Founder Radar Networks .


30 views
Uploaded on:
Description
Making Sense of the Semantic Web. Nova Spivack CEO & Founder Radar Networks. About This Talk. Making sense of the semantic sector Making the Semantic Web more useable Future outlook Twine.com Q & A. The social graph just connects people. The semantic graph connects everything….
Transcripts
Slide 1

Comprehending the Semantic Web Nova Spivack CEO & Founder Radar Networks

Slide 2

About This Talk Making feeling of the semantic area Making the Semantic Web more useable Future standpoint Twine.com Q & A

Slide 3

The social diagram just interfaces individuals The semantic chart associates everything… People Companies Emails Places Products Interests Services Web Pages Activities Documents Projects Events Multimedia Groups The Big Opportunity… Better hunt More focused on promotions Smarter joint effort Deeper incorporation Richer substance Better personalization

Slide 4

The third decade of the Web A period in time, not an innovation… Enrich the structure of the Web Improve the nature of inquiry, cooperation, distributed, publicizing Enables applications to end up more coordinated and keen Transform Web from fileserver to database Semantic advancements will assume a key part

Slide 5

The Intelligence is in the Connections Intelligent Web 4.0 Web OS 2020 - 2030 Intelligent individual specialists Web 3.0 Semantic Web Distributed Search SWRL OWL 2010 - 2020 SPARQL Semantic Databases AJAX OpenID Connections between Information Semantic Search Social Web ATOM Widgets RSS RDF Mashups P2P Web 2.0 Office 2.0 Javascript Flash SOAP XML Weblogs Social Media Sharing 2000 - 2010 The Web Java HTML SaaS Social Networking HTTP Directory Portals Wikis VR Keyword Search Lightweight Collaboration Web 1.0 The PC BBS Websites Gopher 1990 - 2000 MacOS SQL MMO\'s Groupware SGML Databases Windows File Servers The Internet PC Era Email IRC 1980 - 1990 FTP USENET PC\'s File Systems Connections between individuals

Slide 6

Beyond the Limits of Keyword Search The Intelligent Web 4.0 Productivity of Search 2020 - 2030 Reasoning The Semantic Web 3.0 Semantic Search 2010 - 2020 The Social Web Natural dialect look Web 2.0 The World Wide Web 2000 - 2010 Tagging Web 1.0 1990 - 2000 Keyword seek The Desktop Directories PC Era 1980 - 1990 Files & Folders Databases Amount of information

Slide 7

A Higher Resolution Web IBM.com Web Site Joe Person Lives in IBM Company Palo Alto City Publisher of Fan of Subscriber to Lives in Employee of Sue Person Jane Person Dave.com RSS Feed Fan of Coldplay Band Friend of Member of Depiction of Design Team Group Married to Source of Member of 123.JPG Photo Dave.com Weblog Bob Person Depiction of Member of Member of Dave Person Stanford Alumnae Group Author of Member of

Slide 8

Five Approaches to Semantics Tagging Statistics Linguistics Semantic Web Artificial Intelligence

Slide 9

Pros Easy for clients to include and read labels Tags are just strings No calculations or ontologies to manage No innovation to learn Cons Easy for clients to include and read labels Tags are just strings No calculations or ontologies to manage No innovation to learn Technorati Del.icio.us Flickr Wikipedia The Tagging Approach

Slide 10

Pros: Pure scientific calculations Massively scaleable Language free Cons: No comprehension of the substance Hard to specialty great inquiries Best to find truly famous things – not great at discovering needles in sheaves Not useful for organized information Google Lucene Autonomy The Statistical Approach

Slide 11

Pros: True dialect understanding Extract information from content Best for scan for specific realities or connections More exact questions Cons: Computationally escalated Difficult to scale Lots of blunders Language-subordinate Powerset Hakia Inxight, Attensity, and others… The Linguistic Approach

Slide 12

Pros: More exact questions Smarter applications with less work Not as computationally concentrated Share & interface information between applications Works for both unstructured and organized information Cons: Lack of apparatuses Difficult to scale Who makes all the metadata? Radar Networks DBpedia Project Metaweb The Semantic Web Approach

Slide 13

Pros: Smart in thin spaces Answer addresses insightfully Reasoning and learning Cons: Computationally concentrated Difficult to scale Extremely difficult to program Does not function admirably outside of tight areas Training takes a considerable measure of work Cycorp The Artificial Intelligence Approach

Slide 14

The Approaches Compared Make the Data Smarter A.I. Semantic Web Linguistics Tagging Statistics Make the product more brilliant

Slide 15

Two Paths to Adding Semantics "Base Up" (Classic) Add semantic metadata to pages and databases everywhere throughout the Web Every Website gets to be semantic Everyone needs to learn RDF/OWL "Beat Down" (Contemporary) Automatically create semantic metadata for vertical spaces Create administrations that give this as an overlay to non-semantic Web Nobody needs to learn RDF/OWL - Alex Iskold

Slide 16

In Practice: Hybrid Approach Works Best Tagging Semantic Web Top-down Statistics Linguistics Bottom-up Artificial insight

Slide 17

The Semantic Web is a Key Enabler Moves the "knowledge" out of utilizations, into the information Data gets to be self-portraying; Meaning of information turns out to be a piece of the information Apps can get to be more quick witted with less work, in light of the fact that the information conveys learning about what it is and how to utilize it Data can be shared and connected all the more effortlessly

Slide 18

User Profiles Web Content Ads & Listings Data Records Apps & Services Open Query Interfaces Open Data Mappings Open Data Records Open Rules Open Ontologies The Semantic Web = Open database layer for the Web

Slide 19

Semantic Web Open Standards RDF – Store information as "triples" OWL – Define frameworks of ideas called "ontologies" Sparql – Query information in RDF SWRL – Define rules GRDDL – Transform information to RDF

Slide 20

Predicate Subject Object RDF "Triples" the subject, which is a RDF URI reference or a clear hub the predicate, which is a RDF URI reference the question, which is a RDF URI reference , an exacting or a clear hub Source: http://www.w3.org/TR/rdf-ideas/#section-triples

Slide 21

Semantic Web Data is Self-Describing Linked Data Ontologies Definition Data Record ID Field 1 Value Field 2 Value Field 3 Value Field 4 Value Definition

Slide 22

RDBMS versus Triplestore Person Table S P O Subject Predicate Object 001 isA Person 001 firstName Jim 001 lastName Wissner 001 hasColleague 002 002 isA Person 002 firstName Nova 002 lastName Spivack 002 hasColleague 003 003 isA Person 003 firstName Chris 003 lastName Jones 003 hasColleague 004 004 isA Person 004 firstName Lew 004 lastName Tucker f_name jim nova chris lew ID 001 002 003 004 l_name wissner spivack jones tucker Colleagues Table SRC-ID 001 002 003 004 TGT-ID 001 002 003 004 001 002 003 004 001 002 003 004 001 002 003 004

Slide 23

Merging Databases in RDF is Easy S P O S P O S P O

Slide 24

IBM.com Web Site Joe Person IBM Company Palo Alto City Lives in Publisher of Fan of Subscriber to Lives in Employee of Sue Person Jane Person Dave.com RSS Feed Coldplay Band Fan of Friend of Member of Design Team Group Depiction of Married to Source of 123.JPG Photo Member of Dave.com Weblog Bob Person Depiction of Member of Dave Person Stanford Alumnae Group Member of Author of Member of The Web IS the Database! Application An Application B

Slide 25

Are RDF/OWL the Only Way to Express Semantics? Different contenders: String labels Taxonomies and controlled vocabularies Microformats Ad hoc [name, value] sets Alternative semantic metadata documentations

Slide 26

One Semantic Web or Many? The answer is… .Both The Semantic Web is a web of semantic networks Each of us may have our own semantic web…

Slide 27

Why has it Taken So Long? The Dream of the Semantic Web has been ease back to arrive The first vision was excessively centered around A.I. Innovations and apparatuses were deficient Needs for open information on the Web were not sufficiently solid Keyword inquiry and labeling were adequate… for some time Lack of end-client confronting executioner applications Lots of misconception to clear up

Slide 28

Crossing the Chasm… Communicating the vision Focus on open information, not A.I. Innovation advance Standards & devices at last developing Needs were not sufficiently solid Keyword look and labeling not as beneficial any longer Apps require better approach to share information Killer applications and substance Several organizations are beginning to open information to the Semantic Web. Before long there will be a considerable measure of information. Advertise Education Show the market what the advantages are

Slide 29

Future Outlook 2007 – 2009 Early-Adoption A couple executioner applications develop Other applications begin to incorporate 2010 – 2020 Mainstream Adoption Semantics generally utilized as a part of Web substance and applications 2020 + Next huge cycle: Reasoning and A.I. The Intelligent Web The Web learns and thinks on the whole

Slide 30

The Future of the Platform… 1980\'s - The desktop is the stage 1990\'s - The program is the stage 2000\'s - The Web is the stage 2010\'s - The Graph is the stage 2020\'s - The system is the stage 2030\'s - The body is the stage… ?

Slide 31

A Mainstream Application of the Semantic Web…

Slide 32

What is Twine? Twine is another administration for overseeing & sharing data on the Web Works for substance, learning, information, or some other sorts of data Designed for people and gatherings that need a superior approach to arrange, pursuit, share and monitor their data

Slide 33

How Twine Works Collect or creator organized or unstructured data into Twine by means of email, the Web or the desktop Twine makes an information web naturally Understands, labels & joins data consequently Automatically furthers inquire about for you on the Web Organizes data consequently Provides semantic inquiry, disclosure & enthusiasm following Helps you associate with other individuals & gatherings to develop and share information networks around regular premiums

Slide 34

Use-Cases Individuals Collect & creator data about premiums Share with your frien

Recommended
View more...