From RDFaWiki

Contents 1 RDF/XML 2 Comment-Embedded JSON/XML/key-value pairs 3 Microformats 4 Screen Scraping 5 Imperative APIs over HTTP 6 Natural Language Processing 7 Automated Pattern Detection 8 GRDDL-like User-shared Transformations 9 Semantic Style Sheets 10 Site-embedded Data UIs

RDF/XML

• Issues:
  • Requires web page authors to repeat data in two separate documents.
  • Increases the chance of data updates de-synchronizing meaning.

Comment-Embedded JSON/XML/key-value pairs

• Issues:
  • Requires web page authors to repeat data in a page twice
  • Separates the data model from the data on the page - easy for the data model and the data on the page to get out of sync.
  • No unified data model for expressing statements.

Microformats

• Issues:
  • New parser must be written for new Microformats, slowing developer adoption and raising barrier to entry for non-developers.
  • Doesn't solve long-tail problems for people that want to do chemical markup on the web, or markup of species or biological information.

Screen Scraping

• Issues:
  • Scrapers must be re-written or updated when the page layout changes.
  • Scrapers must be written specifically for a specific site and single page type

Imperative APIs over HTTP

• Issues:
  • Requires the developer to learn a new API, even though each service publishes the same type of data (geo-data for example)
  • No way for web browser to auto-discover the data while browsing the web.

Natural Language Processing

• Issues:
  • Remains an "ivory tower" technology, no way for regular web developers to help further the technology.
  • Difficult to apply to general problems and the web-at-large.
  • Does not support distributed innovation.
  • Technology doesn't work for languages that do not have a large number of speakers, punishes people that don't speak English.

Automated Pattern Detection

• Issues:
  • Remains an "ivory tower" technology, no way for regular web developers to help further the technology.
  • Difficult to apply to general problems and the web-at-large.
  • Does not support distributed innovation.
  • Applications for the technology at the current implementation level are niche.

GRDDL-like User-shared Transformations

• Issues:
  • No standard mechanism for doing so.
  • There has been no large community behind such an effort.
  • Lack of corporate support for the technology.
  • Does not actually solve the problem of a standard data model and semantics language.

Semantic Style Sheets

• Issues:
  • Theoretical with no known implementations or definition of the technology

Site-embedded Data UIs

• Issues:
  • Site-specific solution where site owner must have expertise in building semantic expression systems.
  • Does not actually solve the problem of a standard data model and semantics language.