UBhave framework


Published on

UBhave framework part 2 (presentation at MCSS workshop, Zurich, Sept 2013)

Published in: Technology, Education
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

UBhave framework

  1. 1. Ubhave framework architectureIntervention authoring: from social to computer science Intervention distribution: JSON file with BCI definition BCIs at work: context-aware interventions Data collection and transfer: sensing & user querying
  2. 2. Ubhave mobile application Triggers ActivitiesBackground service M V C M V C M V C M V C Dynamic Application Survey Diary Information Settings Sensors Intervention Model App Controller Menu Views Intelligent triggering Data Manager ML
  3. 3. Intervention Task SchedulingMotivation: • Timely delivery of interventions to the user • Information should come when the user is the right environment to accept the message • Contextual information collection with mobile sensing
  4. 4. Intervention Task Scheduling ExamplesIntervention delivery: • Information about available transport options, along with their environmental impact and calories that a user will spend is shown whenever the user leaves home
  5. 5. Intervention Task Scheduling ExamplesContext collection: • Application records presence of other people whenever the user is jogging, eventually identifying people with whom the user exercises the most
  6. 6. Intervention Task Scheduling MethodIntelligent triggering: • Application tasks are defined and executed according to the rules set by the developer May I suggest some healthy options from the menu? Rules: location==restaurant Tasks: notify about healthy foods
  7. 7. Intelligent Trigger JSON file supplied with the application: • Rules: • Time conditioned • Context conditioned: • Location • Activity • Company • Sound • Emotion • Tasks: • Intervention information display • User feedback collection • Sensor data collection
  8. 8. Intelligent Trigger Architecture Android library on top of Sensing, Triggering and Machine learning libraries Triggers Background Sensors Intelligent triggering ML Sample sensors relevant to the prescribed rules Set and monitor alarms Extract high level context from sensor data Manage triggers Check if conditions are met Call appropriate tasks
  9. 9. DEMO
  10. 10. Conclusions Ubhave framework enables: • Digital behaviour change intervention authoring and delivery • Collecting information about the users (sensing) and from the users (surveys, diaries)
  11. 11. Future work Ubhave Intervention framework Integration with large scale BCI management tools Authoring tool for automated BCI design Intelligent triggering library applications Deployment of the framework and BCI applications
  12. 12. Ubhave framework team University of Southampton • Charlie Hargood, Danius Michaelides, Mark Weal University of Cambridge • Neal Lathia, Kiran Rachuri, Cecilia Mascolo University of Birmingham • Veljko Pejovic, Mirco Musolesi
  13. 13. Thank you! ubhave.org Neal Lathia neal.lathia@cl.cam.ac.uk Veljko Pejovic v.pejovic@cs.bham.ac.uk