Detecting Communication Situations Using Smartphone Sensors - Toward Real SNS - Toward

Connections between people are important in social life. The connections are usually recorded using business cards in business situations. However, opportunities for encounters to create a connection are rarely recorded, if people do not take the trouble of recording them. This is especially true concerning information regarding the place or situation that afforded the connection opportunity. That makes it difficult to reuse the connection information. Much research has been conducted so far on automatically detecting situations of a connection between two people. However, this research usually relies on special sensors or tools to recognize the connection. Such sensors or tools are unfortunately not usually readily available in daily life, and people do not want to carry additional sensors on a daily basis if they are only specialized to serve the purpose of recognizing connections. This paper proposes a method to detect situations in the connections between people using the BLE functions, acceleration sensors, microphone and recorder built into an ordinary smartphone. The objective in this research is to make clear the possibilities and the limitations of the proposed method in detecting peoples connection situations. Experimental results suggest that the proposed method can detect the conversation situation between people with their smartphones in their pocket or bag who are standing within a short distance of each other of less than one meter.