Live Topic Generationfrom Event StreamsVuk Milicic, José Luis Redondo Garcia,Giuseppe Rizzo, Raphaël Troncy, Thomas Steine...
Media Finder (www2013)15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 2
Media Finder (zooming on media items)15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 3
Media Finder (timeline view)15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 4
Media Finder (timeline view)15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 5
Media Server Composition of media item extractors (12 SNs) Rely on search APIs + a fix 30s timeout window to provide res...
15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 7Deep linkPermalinkClean text for NLPprocessingAggregat...
Media Finder Architecture Media items harvesting using the Media Serverhttp://eventmedia.eurecom.fr/media-server/search/...
Named Entities are Pivotalhttp://nerd.eurecom.fr/15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 9REST ...
15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 10
15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 11
15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 12
Media Finder (named entities clustering)15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 13
Media Finder (zooming in a cluster)15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 14
Summary Pick an event identified with a hashtag Use MediaServer to get media itemsaggregated over multiple social netwo...
Live Topic Generation from Event Streams Meet us at WWW 2013 Demo Session, Booth 14http://www.youtube.com/watch?v=8iRiwz7...
http://www.slideshare.net/troncy15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 17
Upcoming SlideShare
Loading in …5
×

Live topic generation from event streams

1,257 views

Published on

"Live Topic Generation from Event Streams", talk given at the Demo session of the 22nd World Wide Web Conference (WWW), Rio de Janeiro, Brazil

Published in: Technology, Business
0 Comments
1 Like
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total views
1,257
On SlideShare
0
From Embeds
0
Number of Embeds
193
Actions
Shares
0
Downloads
16
Comments
0
Likes
1
Embeds 0
No embeds

No notes for slide

Live topic generation from event streams

  1. 1. Live Topic Generationfrom Event StreamsVuk Milicic, José Luis Redondo Garcia,Giuseppe Rizzo, Raphaël Troncy, Thomas Steinerraphael.troncy@eurecom.fr / @rtroncy
  2. 2. Media Finder (www2013)15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 2
  3. 3. Media Finder (zooming on media items)15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 3
  4. 4. Media Finder (timeline view)15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 4
  5. 5. Media Finder (timeline view)15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 5
  6. 6. Media Server Composition of media item extractors (12 SNs) Rely on search APIs + a fix 30s timeout window to provide results Fallback on screen scraping when necessary (Twitter ecosystem) Implemented as a NodeJS server Serialize results in a common schema (JSON)22nd World Wide Web Conference (WWW) - Rio de Janeiro15/05/2013 - 6
  7. 7. 15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 7Deep linkPermalinkClean text for NLPprocessingAggregate view of ALLsocial interactions12 Social Networks
  8. 8. Media Finder Architecture Media items harvesting using the Media Serverhttp://eventmedia.eurecom.fr/media-server/search/{combined}/{term}https://github.com/vuknje/media-server (@tomayac fork) Image near de-duplicationDCT signature on image and video frame,Hamming distance between image pairs Clustering and disambiguationNamed Entity Extraction using NERDTopic Generation using LDADensity-based clustering using OPTICS15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 8
  9. 9. Named Entities are Pivotalhttp://nerd.eurecom.fr/15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 9REST API OntologyDashboard UI
  10. 10. 15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 10
  11. 11. 15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 11
  12. 12. 15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 12
  13. 13. Media Finder (named entities clustering)15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 13
  14. 14. Media Finder (zooming in a cluster)15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 14
  15. 15. Summary Pick an event identified with a hashtag Use MediaServer to get media itemsaggregated over multiple social networks Use NERD to get entitiesaggregated over multiple extractors Cluster and identify meaningful topics(aka entities)with a meaningful labeloften disambiguated with a DBpedia URI giving accessto more encyclopedic knowledge15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 15
  16. 16. Live Topic Generation from Event Streams Meet us at WWW 2013 Demo Session, Booth 14http://www.youtube.com/watch?v=8iRiwz7cDYY15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 16
  17. 17. http://www.slideshare.net/troncy15/05/2013 22nd World Wide Web Conference (WWW) - Rio de Janeiro - 17

×