A framework for context-aware adaptation in public displays

Presentation at the CAMS 2009 OTM workshop in Algarve, Portugal.

1. A Framework for Context-Aware Adaptation in Public Displays Jorge C. S. Cardoso jorgecardoso@ieee.org Rui José rui@dsi.uminho.pt DSI, Universidade do Minho, Guimarães
2. Public displays For public displays “the right information at the right time”, depends on place, i.e., people + location + time
3. Public displays ...displays need to be able to adapt to the (short and long term) activity of people nearby how can displays sense this activity?
4. Footprints We can tell something about the person by looking at a footprint, e.g., Barefeet / Shoes on Man / Woman
5. Not all footprints are the same... ...some, may give us interesting hints: Modern, urban, fashion Sports
6. Not all footprints are the same... ...other, don’t tell us much, or are misleading...
7. Digital footprints for public displays So, what footprints are there for public digital displays?
8. Identifying digital footprints Identified footprints that result from interacting with a public digital display. Analysed display applications and types of interactive features supported A footprint must contribute something to the display’s knowledge about it’s social environment
9. Digital footprints Footprints create the display’s (mostly social) context.
10. Presence detection Presence detection – The display detects the presence of nearby people but is only able to determine that someone is around Can be achieved with • proximity sensors • basic computer vision techniques Can be used by the display to • trigger eye-catching content • switch between ambient and interactive modes
11. Presence detection - example Ju, W.; Lee, B. A. & Klemmer, S. R. Range: exploring implicit interaction through electronic whiteboard design CSCW '08: Proceedings of the ACM 2008 conference on Computer supported cooperative work, ACM, 2008, 17-26 Range Whiteboard
12. Presence characterisation Presence characterisation – The display detects the presence of nearby people and is able to determine how many, estimate their gender and age what they are look at, etc Can be achieved with • computer vision techniques • people counter sensors Can be used by the display to • adjust content to audience type (male/female) • determine interest
13. Presence characterisation - example TruMedia Technologies: http://www.trumedia.co.il/ Trumedia Proactive Advertising Estimates •  Viewers •  Gender •  Age Selects ads based on rules
14. Presence characterisation - example Sawhney, N.; Wheeler, S. & Schmandt, C. Aware Community Portals: Shared Information Appliances for Transitional Spaces Personal Ubiquitous Computing, Springer-Verlag, 2001, 5, 66-70 Aware Community Portals Infers interest based on time looking at display Creates an implicit rating for Slashdot articles
15. Presence identification Presence identification – The display detects the presence of nearby people and is able to identify them (i.e., relate the presence of the same person on different ocasions) Can be achieved with • RFID, Bluetooth, Magnetic cards, “manual logins” Can be used by the display to • prevent repetition of content • combine other information and provide personalized content
16. Presence identification - example Sharifi, M.; Payne, T. & David, E. Public Display Advertising Based on Bluetooth Device Presence Mobile Interaction with the Real World (MIRW 2006) in conjunction with the 8th Intl Conference on Human Computer Interaction with Mobile Devices and Services, 2006 BluScreen Detects Bluetooth devices Selects an advert that most people have not yet seen
17. Presence identification - example McCarthy, J. F.; Farnham, S. D.; Patel, Y.; Ahuja, S.; Norman, D.; Hazlewood, W. R. & Lind, J. Supporting community in third places with situated social software C&T '09: Proceedings of the fourth international conference on Communities and technologies, ACM, 2009, 225-234 CoCollage Loyalty card or web page login to indicate presence Shows content items submitted from users and prefers items from present users
18. Self-exposure Self-exposure − User tells something about himself Can be achieved with • web profiles • custom mobile applications • bluetooth naming Can be used by the display to • adjust to individual preferences
19. Self-exposure - example McDonald, D. W.; McCarthy, J. F.; Soroczak, S.; Nguyen, D. H. & Rashid, A. M. Proactive displays: Supporting awareness in fluid social environments ACM Trans. Comput.-Hum. Interact., ACM, 2008, 14, 1-31 Proactive Displays Combined with identification (RFID) Users fill in a web profile The displays react to user’s presence in different ways: • AutoSpeakerID: shows name and affiliation of person asking a question during the Q&A time • Ticket2talk: shows personal interests of people near the display in the “coffee break” area • NeighborhoodWindow: shows common keywords to the group near the display
20. Self-exposure - example José, R.; Otero, N.; Izadi, S. & Harper, R. Instant Places: Using Bluetooth for Situated Interaction in Public Displays Pervasive Computing, IEEE, 2008, 7, 52-57 Instant Places Combined with identification (Bluetooth) Users use custom profile commands in Bluetooth device name
21. Content suggestion Content suggestion – User sugests content to display Can be achieved with • email • OBEX • web Can be used by the display to • display similar content • infer user interests
22. Content suggestion - example Churchill, E. F.; Nelson, L.; Denoue, L.; Helfman, J. & Murphy, P. Sharing multimedia content with interactive public displays: a case study DIS '04: Proceedings of the 5th conference on Designing interactive systems, ACM, 2004, 7-16 Plasma Poster Network
23. Actionables Actionables – User acts in response to display Download – Dowloads an item Control Content – Exerts control over the display of items (stop, display next, etc) Rate – Rates an item Vote – Votes on a poll Can be used by the display to • indirectly determine user preferences
24. Actionables - example Grasso, A.; Muehlenbrock, M.; Roulland, F. & Snowdon, D. O'Hara, K.; Perry, E.; Churchill, E. & Russel, D. M. (ed.) Supporting communities of practice with large screen displays Public and Situated Displays - Social and Interactional Aspects of Shared Display Technologies, Kluwer, 2003, 261-282 Community Wall Users can rate items Rating is used in scheduling algorithm to determine which items to show
25. Conclusion So what? The digital footprints set the (social) context for public display systems.
26. Conclusion This set of digital footprints can serve as a framework developing adaptable public display systems. Provide a mapping between interactive features, relevant information for the display and adaptation strategies Digital footprints provide a way for the display to infer and characterise its social environment
27. Thank you! This presentation is also on http://slideshare.net/jorgecardoso (tag: cams09) A Framework for Context-Aware Adaptation in Public Displays Jorge C. S. Cardoso jorgecardoso@ieee.org Rui José rui@dsi.uminho.pt