My research interests are mainly in spatio-temporal data analysis. I am currently working on data collected in big cities and exploring the patterns affecting the development of urban areas, as well as the effect of urban social interactions on the risk of privacy.
I am also interested in crowd-sourcing methods. In this direction, I recently collaborated in studies on the power of friendships in peer-pressure mechanisms, as well as the effect of social relations in the effectiveness of wisdom of the crowd mechanisms.
Quantitative Land Use Planning
The recent buzz on smart cities has started to revolutionize how cities are planned and built. One of the processes of city planning that did not yet capitalize on the data revolution is land use planning: the task of planning how different types of activities (commercial or residential) are distributed across a city. Together with Talia Kauffman and Erez Shmueli, we are studying a way to introduce quantitative methods in the daily practice of land use planning.
Smartphones connect us now more than ever. We are more and more used to have connectivity on-the-move and the network of free Wi-Fi hotspots keep increasing. Together with Yves-Alexandre de Montjoye, we are studying a way to quantify the privacy risks rising from this increased connectivity in a urban scale.
Friendship as a drive for behavioral change
Social pressure can greatly support behavioral change when an adequate mechanism is in place. We found that when the effect of social pressure is even higher when it involves friends that share a reciprocal tie: meaning that both consider the other a friend. Also, we conducted an empirical study that showed how people are bad judges of friendships: we think that most of our friendships are reciprocal, while only roughly half actually are. [“Are You Your Friends’ Friend? Poor Perception of Friendship Ties Limits the Ability to Promote Behavioral Change” [link], Abdullah Almaatouq, Laura Radaelli, Alex “Sandy” Pentland, Erez Shmueli, PloS one, 2016]
We are studying a solution for exploiting the potential of ride-sharing by making it attractive for all the involved players. We start by studying how to predict the potential benefits of a ride-sharing solution for a specific urban network. This will allow us to make policy makers and city administrators aware of the potential of ride-sharing and willing to promote it through policies and direct incentives. [“Ride sharing: a network perspective” [link], Erez Shmueli, Itzik Mazeh, Laura Radaelli, Alex “Sandy” Pentland, Yaniv Altshuler, SBP 2015]
Unicity of Behavioral Metadata
We investigated how sharing behavioral metadata datasets can compromise anonymity of a user. I worked on this project in collaboration with Yves-Alexandre de Montjoye and Vivek K. Singh during my visit at the Human Dynamics group at MIT MediaLab (Sept 2013-Feb 2014). [“Unique in the shopping mall: On the reidentifiability of credit card metadata” [link], Yves-Alexandre de Montjoye, Laura Radaelli, Vivek Kumar Singh, and Alex “Sandy” Pentland, Science, vol. 347 no. 6221, 30 January 2015]
The focus of my PhD studies was on indoor trajectory data, indoor positioning systems, and indoor mobile services. I am still interested in analyzing indoor movement data and investigating new indoor positioning systems, but I broadened the scope of my research to the outside world.
- Video-Assisted Wi-Fi Indoor Positioning: Together with Yael Moses (IDC Herzliya), we explored the integration of different sources for indoor positioning, namely, Wi-Fi and fixed cameras. [“Towards Fully Organic Indoor Positioning” [link], Laura Radaelli and Christian S. Jensen, ISA@SIGSPATIAL 2013; “Using Cameras To Improve Wi-Fi Based Indoor Positioning” [link] [pdf], Laura Radaelli, Yael Moses and Christian S. Jensen, W2GIS 2014]
- Indoor Movement Pattern Mining: We developed a method that finds frequent paths used by people in indoor spaces. We are looking into ways of extending the method to integrate temporal dimension and of identifying similar (but not exact matching) patterns. [“Identifying Typical Movements Among Indoor Objects‒Concepts and Empirical Study”[link], Laura Radaelli, Dovydas Sabonis, Hua Lu, and Christian S. Jensen, MDM 2013]
- Indoor Activity Recognition: For my Master’s thesis at the EverywareLab@University of Milan, we experimentally evaluate the effectiveness of the ontological approach to activity recognition, using a dataset collected in a smart-home setting. [“Is ontology-based activity recognition really effective?” [link], Daniele Riboni, Linda Pareschi, Laura Radaelli, and Claudio Bettini, CoMoRea@PerCom 2011]