Nearness sensing and interest traces

performed with psychology students, in two different institutions, where the purpose was to study interest influence in psychological proximity. NSense has been installed in Android smartphones carried by a population of 50 students of the two different universities, numbered User1 to User50. The students carried NSense around during their daily routines, for 2 days: 05.04.2017-06.04.2017. The readings obtained show a total of 15 students out of the original universe of 50. The traces collected comprise the following fields: - date: DD/MM HH:mm - own_device_name: unique identifier of the device, Userx - connected_device_name: peer observed via Wi-Fi Direct (range of 0 to 100 meters) at date. - latitude, longitude: GPS coordinates of the device - distance: relative distance computed in meters via Wi-Fi Direct. - sound: discrete value for the surrounding sound activity: ALERT, NORMAL, QUIET - physical activity: discrete value for the type of activity (MOVING or STATIONARY) - tct: total contact time. Starts counting after 1 day of use. Corresponds to the sum of contact duration between i and j on a time window of duration h. - social_strength_minute: level of social interaction derived from our work NSense: A People-centric, non-intrusive Opportunistic Sensing Tool for Contextualizing Social Interaction, IEEE Healthcom2016.Social strength of node i towards node j, in a specific hourly sample h, for day d. Node i computes the social strength towards j by adding the different weighted ADs. t has been set for 24 hours - si: Social interaction of node i towards j, computed at instant t - p: propinquity: measures the probability of social interaction occurring over time. - ema_cd. Exponential moving average of the total contact duration between nodes i and j - additional columns: types of interests defined in NSense.