A total of 35 students participated during spring and summer semesters of 2015.
During two field session weeks, we collected data from 35 participants' everyday lives.
RU Sensor Study app, which is based on funf framework, captured participants' phone call/SMS history as well as GPS coordinates. All data collected through the app was anonymized.
Fitbit HR health-related behavioral signals from the users: number of steps, time of working out, time of sleep, etc.
We asked participants to conduct exploratory searches to write reports about given topics.
|#||Task Type||Task Topic|
|1||Individual Task||Data Security|
|2||Collaborative Task||Health and Wellness|
Probing the Interconnections Between Geo-Exploration and Information Exploration Behavior
Abstract: As increasingly diverse facets of human life - including socializing exercising, and information-seeking - are mediated by ubiquitous technology, they open the doors for the study of the hitherto underexplored interconnections between them. This work motivates and grounds the use of geo-exploration data to predict the information exploration behavior of users and to support their search. Based on a two-week field study involving 35 participants, we have identified multiple geo-exploration features that have significant associations with a user’s information exploration behavior. We also found that the same geo-exploration features could be combined to build predictive models for various facets of an individual’s information exploration behavior, and these models performed significantly better than comparable personality-based models.
Which Team Benefits from Collaboration?: Investigating Collaborative Information Seeking Using Personal and Social Contextual Signals
Abstract: Collaboration often involves looking for information together to achieve a common goal, such as a collaborative search task. A general expectation in collaboration is that through working together, participants obtain better outcomes than those achieved by individual efforts. However, this is not always true and certain collaborative teams benefit more from collaboration than others. This study aims to identify the settings in which teams perform the best in exploratory search tasks. Specifically, this study considers a variety of social signals to characterize individuals (e.g. number of phone calls, number of SMS messages) and identify the pairs of individuals who are likely to perform best in collaborative information seeking tasks. Based on an exploratory multi-method study (N=35) involving “in-the-wild” phone data collection for two weeks and one in-lab search session, we report that: (1) the difference in social activities between the collaborators' daily lives was found to be associated with collaborative information seeking performance and (2) one member of a team – one who has lower social interaction – might significantly influence the outcomes for the team. The results pave way for future study design and analysis in the area of sensor-enhanced collaborative information seeking.