1 / 42

U C L

Inter-Context Trust Bootstrapping for Mobile Content Sharing (daniele quercia) (stephen hailes & licia capra). U C L. What do I do?. Research @. what I research?. Reputation Systems for Mobiles. What’s that?. Example: antique markets. Problem: Visitors cannot see prices of everything!.

paki-garrett
Télécharger la présentation

U C L

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Inter-Context Trust Bootstrappingfor Mobile Content Sharing(daniele quercia)(stephen hailes & licia capra) UCL

  2. What do I do?

  3. Research @

  4. what I research?

  5. Reputation Systems for Mobiles

  6. What’s that?

  7. Example:antique markets

  8. Problem:Visitors cannotsee prices of everything!

  9. Solution:Sellers disseminate e-ads, and visitors collect them

  10. Problem: Sellers may disseminate irrelevant ads

  11. Proposal:

  12. They may keep track of which sellers sendirrelevant ads

  13. Daniele Quercia Trust model on A:how A decideswhether to rely on Bto visit a stall

  14. Daniele Quercia To decide whether to rely on B, A has to set its initial trust in B

  15. 3 Existing Solutions Daniele Quercia

  16. Daniele Quercia • 1. Fixed values • ( over-simplified)

  17. Daniele Quercia • 2. Recommendations • ( fake ones)

  18. 3. Similar contexts • ( universal ontology) Daniele Quercia

  19. Daniele Quercia Two cases: B is 1. unknown 2. partly known

  20. Daniele Quercia • 1. B is unknown

  21. Daniele Quercia • Popular way: • Trust propagation (transitivity) C ? A B

  22. Daniele Quercia • Meant for the Web & Proved on “binary” ratings

  23. Algorithm rating • unrated trust relationships (needed) Daniele Quercia • unrated nodes (chosen) ? AB AC CB 1 2 C 1 2 ? A B

  24. ? • Idea: • 1. Similar nodes together • 2. Find function: • same ratings for rated nodes • similar ratings for neighbours

  25. Daniele Quercia • Tested on real data • (Advogato: > 55K user ratings)

  26. Daniele Quercia • 2. B is partly known

  27. Daniele Quercia Popular way: Inter-context Lifting  Antiques Coins Chairs Roman Coins Greek Coins

  28. Idea: Users … • > Don’t share ontology • > Extract “features” • from their own ratings Daniele Quercia

  29. Idea: Users … • > Don’t share ontology • > Extract “features” • from their own ratings Daniele Quercia

  30. Daniele Quercia • How to extract?

  31. Daniele Quercia • Singular • Value • Decomposition

  32. Beauty: features • not user-specified • BUT learnt Daniele Quercia

  33. Tested on simulation with real parameters Daniele Quercia

  34. Daniele Quercia • Tested on Nokia 3230 • Max: 3.2 ms !

  35. Daniele Quercia • What I’ve told you is on • “mobblog UCL” (google it) • under tag: “bootstrapping”

  36. Daniele Quercia

  37. Daniele Quercia • And User Privacy?

  38. Daniele Quercia • Private filtering • (Google for “mobblog private filtering”)

  39. Daniele Quercia • And Resource Discovery?

  40. Daniele Quercia • Folksonomy for mobiles 

  41. Daniele Quercia • And Attacks?

  42. Daniele Quercia Further Research (join mobblog !)

More Related