• <More on Intel.com

Context Awareness Social Proximity Detection: Project

By monitoring social interactions, a computing device can provide input for contextually aware applications. For example:

  • A smartphone can offer restaurant recommendations based on the people you are with, instead of considering only your preferences. 
  • A device can tag photos of the people who were with you at an event. This simplifies the process of sharing these photos with the group or determining which photos to show based on who is present. 
  • A context-aware computing device can help you with introspection by monitoring your social interaction patterns. It might, for example, remind you to connect with a friend with whom you had been out of touch.
  • Monitoring social interactions can provide high-level insights into how social activities affect your life and well being.

This social-proximity detection research focuses on inferring social interactions. Devices serve as proxies for their users. By monitoring which devices are “within conversational distance” of one another, social interaction can be inferred. To accurately detect conversational distance, data collected from several different sources are aggregated.

  • First, a coarse measure of physical distance identifies possible devices within proximity, using GPS, Wi-Fi*, and Bluetooth*. 
  • Audio proximity further refines the proximity detection by correlating audio characteristics from the different devices. Generally, this approach more accurately reflects social interactions than simple physical distance: Audio is a better means for comprehending barrier and noise effects.
  • Finally, to provide a complete view of social interactions, the use of chat and social networking applications is monitored to infer virtual interactions.

The data collected as a part of social-proximity detection can be effectively combined with data from other types of context-aware computing inputs to enhance the overall user experience. 

Who am I with? Social Proximity Classification

Research Agenda

  • Identify areas where social proximity recognition data can be fused with other forms of context-aware information to create innovative mobile-computing applications.
  • Develop algorithms using context-aware principles and methods that can efficiently be implemented on today’s ultra mobile computing devices
  • Discover techniques to improve the responsiveness and usefulness of computers in supporting the lifestyles and activities of computer users

Recent Publications

Cook, Diane J., Aaron Crandall, Geetika Singla, and Brian Thomas. Detection of Social Interaction in Smart Spaces. Cybernetics and Systems, 41(2): 90–104, 2010.

Learn More

Overview of a wearable brainwave monitor to improve patient care >

Technology to personalize personal training >

Context Awareness to Radically Change How We Interact with Technology >

Researchers Invent the Future at Annual R&D Event >