Heckling at Ontologies

Heckling at Ontologies: bottom-up vs. top-down approaches to metadata generation

Heckling at Ontologies is a collaboration between MAT researchers Toby Harris and Saul Albert exploring the dissonance between top-down and bottom-up approaches to media metadata generation.

The project started with two datasets:

A semantic description of Season 4, Episode 1 of Doctor Who

This was one of the outcomes of Toby’s industrial placement at the BBC, which developed an ontology for rich description of BBC programme content and then used it to create a highly detailed summary of a specific episode of Doctor Who.

User-generated tweets ‘heckled’ at the screen by viewers

As part of Saul’s industrial placement at BT, he worked with The People Speak to develop Heckle, a web service that gathers tweets, images, videos and texts ‘heckled’ to a shared screen by viewers of a TV show, building up a reviewable transcript of the conversation.

These two data sets were analysed comparatively, to discover the dissonances and regularities between Toby’s top-down and Saul’s bottom-up descriptions of the same TV show.

They are building a demo that shows the two data sets superimposed on a live stream of the video. You can download the source of the demo and grab copics of semantic data on the ontoheckle github page.

This entry was posted on Tuesday, November 15th, 2011 at 4:08 pm. It is filed under links, showcase and tagged with html5, metadata, ontologies, semantic web, semweb, socialtv, TV, video. You can follow any responses to this entry through the RSS 2.0 feed.