Dynamic Semantic Publishing hosts the Olympics at the BBC

Starting with the 2010 World Cup, the BBC has been generating sports pages via Dynamic Semantic Publishing. In 2012 the BBC went further and used Dynamic Semantic Publishing to create the Olympics website.

In a Dynamic Semantic Publishing architecture, new content submitted by journalists is automatically analysed (text analysis) against an vocabulary of known players, teams, terminology and so-on and annotated according to the content of the story. When a user requests a page for a team, player or event the system dynamically assembles the page from the appropriate content elements annotated with the desired player (or team, or event…)

This combination of automated indexing and annotation of new content, and dynamic page generation significantly reduces the effort required to update and maintain a website, allowing you to focus your effort on creating valuable new content.

With, as Jem Rayfield put it, “Far too many web pages for far too few journalists”, a conventional approach by the BBC would have required an editor to maintain each index with the latest stories. The simplification of workflow afforded by automated content ingestion and automated dynamic page generation made the BBC’s Sports projects project possible.

The numbers tell the scale of the task:

The World Cup site consists of:

  • Over 750 dynamically generated pages
  • serving over 2,000,000 requests/day
  • accepting 100s of updates/minute

The Olympics site is far larger:

  • 10,000+ Athlete pages
  • 200+ Country pages
  • 500 Event & Discipline pages
  • Dozens of Venue pages
  • Near real-time updates across the competition

At more than 10,000 pages, the job of producing a real-time updated Olympics website was beyond a conventional static CMS stack and required the Dynamic Semantic Publishing approach to make it possible.

See also: