(4) Problem of Semantics: Syntax vs. Meaning

• Computation is great at syntax and grammar
  • All X are Y
• Take the following syllogisms
  • All dogs are mammals
  • All cats are bipeds
  • All Lady Gagas are cats
• Logical syntax does not necessarily take into consideration meaning
  • 'The sum of the angles of a triangle is equal to the color red'
  • While grammatically correct this makes no sense

(5) Adding Meaning to HTML

• HTML describes structures in a very general way
  • HTML elements describe logical page structures such as headings, lists, tables, …
  • useful for dynamic and adaptive page rendering, but not for understanding contents
• Good HTML may have more information available
  • classes in HTML elements may represent underlying concepts (CSS may use this)
  • HTML containers (e.g. lists) may represent aggregation of some basic information items
• Very good HTML
  • standard class names and some guidelines for using them
  • some rules for using combinations of particular class names

(6) Plain HTML

<div>
  <h1>Spoon, Deerhunter, and Micachu and the Shapes at The Fox Theater</h1>
  <p>Tuesday 13 April 2010 at 7:30pm</p>
  <div>
    <div>The Fox Theater</div>
    <div>
      1807 Telegraph Avenue<br/>
      Oakland, CA 94612
    </div>
    <div>(510) 302-2277</div>
  </div>
</div>

(7) Good HTML

<div class="event">
  <h1 class="title">Spoon, Deerhunter, and Micachu and the Shapes at The Fox Theater</h1>
  <p class="when">Tuesday 13 April 2010 at 7:30pm</p>
  <div class="where">
    <div class="venue">The Fox Theater</div>
    <div class="address">
      <div class="street-address">1807 Telegraph Avenue</div>
      <span class="city">Oakland</span>, 
      <span class="state">CA</span>
      <span class="zip">94612</span>
    </div>
    <div class="phone">(510) 302-2277</div>
  </div>
</div>

(8) Very Good HTML: Microformats

<div class="vevent">
  <h1 class="summary">Spoon, Deerhunter, and Micachu and the Shapes at The Fox Theater</h1>
  <p class="date">
    <abbr title="20100413T1930Z" class="dtstart">Tuesday 13 April 2010</abbr> 
    at <span>7:30pm</span>
  </p>
  <div class="location vcard">
    <div class="fn org">The Fox Theater</div>
    <div class="adr">
      <div class="street-address">1807 Telegraph Avenue</div>
      <span class="locality">Oakland</span>, 
      <abbr class="region" title="California">CA</abbr> 
      <span class="postal-code">94612</span>
    </div>
    <div class="tel">(510) 302-2277</div>
  </div>
</div>

(9) Microformats

• A microformat is an agreed-upon set of names for HTML element classes, and rules for using them
• Microformats don't add anything new to HTML: they're a way of using HTML
• Microformats are good for small, commonly encountered bits of data
  • Calendar events [http://microformats.org/wiki/hcalendar], contact info [http://microformats.org/wiki/hcard], licensing info [http://microformats.org/wiki/rel-license], etc.
  • nofollow [Web Search; nofollow (1)] is a microformat too
• Creating new microformats involves reaching consensus in the microformat community
  • This takes time and effort, so there are lots of "draft" microformats and few actually in use
  • Only the most common kinds of data get standardized

(10) Very Good HTML: RDFa

<div class="vevent">
  <h1 class="summary">
    <span about="http://dbpedia.org/resource/Spoon_%28band%29"
          typeof="http://purl.org/ontology/mo/MusicGroup">
    <span property="http://xmlns.com/foaf/0.1/name">Spoon</span></span>, 
    <span about="http://dbpedia.org/resource/Deerhunter"
          typeof="http://purl.org/ontology/mo/MusicGroup">
      <span property="http://xmlns.com/foaf/0.1/name">Deerhunter</span></span>, 
    and 
    <span about="http://dbpedia.org/resource/Micachu"
          typeof="http://purl.org/ontology/mo/MusicArtist">
      <span property="http://xmlns.com/foaf/0.1/name">Michachu</span></span>, 
    and the Shapes at The Fox Theater
  </h1>
  <p class="date">
    <abbr title="20100413T1930Z" class="dtstart">Tuesday 13 April 2010</abbr> 
    at <span>7:30pm</span>
  </p>
  <div class="location vcard">
    <div class="fn org">The Fox Theater</div>
    <div class="adr">
      <div class="street-address">1807 Telegraph Avenue</div>
      <span class="locality">Oakland</span>, 
      <abbr class="region" title="California">CA</abbr> 
      <span class="postal-code">94612</span>
    </div>
    <div class="tel">(510) 302-2277</div>
  </div>
</div>

(11) RDFa

• RDFa is another approach to adding meaning to HTML
• Unlike microformats, RDFa adds some new attributes to HTML
• While microformats are defined for specific kinds of data, RDFa can be used for any kind of data
  • No need for wide consensus
• But RDFa is a little more complicated to use than microformats

(12) Using Microformats & RDFa with Operator

• Operator [https://addons.mozilla.org/en-US/firefox/addon/4106] is a Firefox add-on for interacting with microformats and RDFa
• It looks for microformat and RDFa classes and attributes and does intelligent things with them
• Its functionality can be extended with user scripts
• Try it on our example [src/event-rdfa.html], or Yelp [http://www.yelp.com/biz/top-dog-berkeley], or Best Buy [http://products.semweb.bestbuy.com/y/products/9811355/]

(14) Tim Berners-Lee Linked Data Vision

(15) Linked Data Principles

1. Use URIs as names for things.
2. Use HTTP URIs so that people can look up those names.
3. When someone looks up a URI, provide useful information, using a standard data model.
4. Include links to other URIs, so that people can discover more things.

(16) URIs as Names

• Recall [URIs & HTTP; Resources & Representations (1)] that a URI identifies an abstract resource
• This means that a URI can identify any abstract concept, not just web pages
• The Semantic Web is mainly about how to assign URIs to any concept
• DBpedia [http://dbpedia.org/] is a project creating URIs for every topic in Wikipedia
  • Spoon: http://dbpedia.org/resource/Spoon_%28band%29 [http://dbpedia.org/resource/Spoon_%28band%29]
  • Mysticism: http://dbpedia.org/resource/Mysticism [http://dbpedia.org/resource/Mysticism]
• But anyone can create URIs for concepts
  • Panthera tigris: http://lod.geospecies.org/ses/QMUrD [http://lod.geospecies.org/ses/QMUrD]
• Multiple people or groups can create URIs for the same concept
  • All the URIs for Panthera tigris [http://sameas.org/html?uri=http%3A%2F%2Flod.geospecies.org%2Fses%2FQMUrD]

(17) Semantic Web Data Model

• The standard data model defined by Semantic Web standards is called RDF
  • Stands for Resource Description Framework
• Data in RDF is just lists of statements:
  • Spoon is a music group
  • Spoon is named "Spoon"
  • Spoon has a member Britt Daniel
• A statement consists of three parts: [subject] [predicate] [object]
  • [Spoon] [is a] [music group]
  • [Spoon] [is named] ["Spoon"]
  • [Spoon] [has a member] [Britt Daniel]

(18) Semantic Web Data Model

(19) Semantic Web Data Model

(20) Linking Data Sets

(22) Enhanced Search Results

• When indexing [Web Search; Indexing (1)] web pages, search engines can parse microformats and RDFa to add details to result pages
• Instead of trying to guess at address, prices, or reviews, they can access this data directly
• Doesn't affect ranking, only the display of results
  • So there's no point in putting in fake/spam data

(23) Practical Applications

[http://search.yahoo.com/search?p=avila+boston]

(24) Practical Applications

[http://www.google.com/search?q=fillmore+new+york+concerts]

(25) Government Data

• Government data sets are a good use case for the Semantic Web
  • Government data is intended to be made available to the public
  • Publishing on the web is a great way to make data widely available
• Data sets are more useful when they can be linked and merged
• Linking and merging before publishing to some centralized place creates a bottleneck
• In theory, the Semantic Web can enable linking and merging of data sets published independently
• Theory being tested at data.gov.uk [http://data.gov.uk/]

(30) Metadata as Markup Overlay

(31) Surfacing Concepts

• [@microformats] are embedded into HTML Web content
  • there is no need for alternative representations
  • machines use the same resources as humans (and extract microformat data)
• Many larger Web sites are using structured back-ends
  • for pure HTML publishing, semantics are translated into HTML/CSS
  • more semantics can be represented by also producing microformats
• Microformats can also help in decentralized service scenarios
  • centralized: data has to be submitted to a hub
  • decentralized: crawlers search for microformats and use what they find
  • advantage of decentralization: loose coupling and independent control

(33) Different Representations

(34) XML as Custom Markup

• XML [../xml-fall10] can represent arbitrary (textual) data
  • works very well for tree-structured and document-style data
  • works not so well for graph-like data with no inherent order
• XML only defines a syntax for representing ordered trees
  • what the elements and attributes mean requires agreement on the vocabulary
  • XML is a good tool for data exchange, but only for the syntax part
• Agreement can be based on three different approaches
  1. everybody always uses a universally useful vocabulary (XHTML)
  2. user (groups) agree on vocabularies based on mutual interest
  3. user (groups) agree on modules and build vocabularies with these modules (UBL [http://www.oasis-open.org/committees/tc_home.php?wg_abbrev=ubl])

(36) Translation into a Universal Metamodel

(37) The Semantic Web Vision

• [@rdf] is a more general model for structured data
  • it is very fine-granular and can represent (almost) everything
  • its granularity becomes problematic in scenario with coarse granularity
• Transforming resources can be done in two basic ways
  1. using well-defined mappings between HTML and RDF (GRDDL [http://www.w3.org/TR/grddl-primer/])
  2. extracting information by analysis such as NLP [http://en.wikipedia.org/wiki/Natural_language_processing]
• Transforming resources does not require cooperation of the service
  • any service can be crawled and the resources can be transformed
• Transforming resources does not allow access to all data
  • crawling can have limits and thus the RDF data is limited as well
  • with everything as RDF the complete RDF graph could be queried

(39) Bypassing Web Publishing

(40) Trillions of Triples

• Turn the structured data from the back-end into RDF
  • the complete dataset can be transformed in one processing step
  • relationships between resources may be better preserved than by crawling
  • keeping the snapshot current can become a significant problem
• With everything in one database (a triple store [http://en.wikipedia.org/wiki/Triplestore]) queries become possible
  • SPARQL queries can query the graph of complete service data
  • large datasets easily translate into several billion triples
  • triple store implementation moves to native RDF databases
• Large triple stores can process complex SPARQL queries
  • performance is a problem with large datasets
  • RDF's simplicity bites back in the form of reification [http://en.wikipedia.org/wiki/Reification_%28computer_science%29#Reification_on_Semantic_Web]

(42) Exposing Structured Data

• Exposing semantics makes data more valuable
• Various technologies and approaches can be chosen
• What are the economics behind allowing/supporting reuse?
• Different implementations fit different scenarios