A System to Analyze Group Socializing Behaviors in Social Parties

• Published in: IEEE transactions on human-machine systems Vol. 47; no. 6; pp. 801 - 813
• Main Authors: Hyukjae Jang, Choe, Sungwon P., Gunkel, Simon N. B., Seungwoo Kang, Junehwa Song
• Format: Journal Article
• Language: English
• Published: IEEE 01-12-2017
• Subjects: Behavioral sciences; Context awareness; Context-aware services; Man-machine systems; Monitoring; networking party; Noise measurement; Real-time systems; Sensors; social interaction; Social network services; socializing behavior; Ubiquitous computing
• ISSN: 2168-2291; 2168-2305
• DOI: 10.1109/THMS.2016.2634918

Social events such as networking parties are excellent opportunities to expand one's social network and make new social ties, as well as simply have fun and enjoy oneself. With the emergence of pervasive technology, we now have the opportunity to discover face-to-face socializing behaviors of individual guests and support them in their socializing goals. Quantitative measures of socializing behaviors will be useful in diverse applications from social therapy to social event management. We propose a system that detects and analyzes group-level socializing behaviors to support real-time and ex post facto social applications deployed in real social event situations. In our approach, we exploit attendees' characteristics of socializing such as proximity and body orientation to determine social proximity among guests and subsequently detect socializing groups. We performed preliminary studies and multiple deployments of our system and example applications in real social networking parties. Evaluation shows the system can detect socializing groups accurately and support applications that satisfy the interests of guests and party hosts.