Popularity Prediction for Social Media over Arbitrary Time Horizons
Daniel Haimovich, Dima Karamshuk, Thomas Leeper, Evgeniy Riabenko, Milan Vojnovic
IDMC Global Report on Internal Displacement (GRID)
In this paper, we present a novel approach for using aggregated and anonymized Facebook location data to measure displacement patterns in the weeks and months after disasters. This methodology has been developed through a collaboration between the Internal Displacement Monitoring Centre (IDMC) and Facebook as part of the Disaster Maps initiative. The displacement maps described here leverage Facebook location-history (LH) data. LH is an opt-in setting in the Facebook app, where people consent to sharing location data in order to enable location-based services (e.g., Nearby Friends, location-based ads) and social-good products like Disaster Maps. Individual-level LH data is sensitive, and misuse could compromise the privacy and safety of individuals and communities. In this paper we explain the privacy protection mechanisms we use in order to protect the safety of Facebook users. This paper will also explore the insights we have found in two specific crises: Cyclone Fani in India and Bangladesh and Typhoon Hagibis in Japan. For Japan we present insights we have obtained after fielding a survey to those that have been displaced by the typhoon. Last, we compare our data to estimates collected by IDMC and explore potential hypotheses that could explain why these estimates are different.
Daniel Haimovich, Dima Karamshuk, Thomas Leeper, Evgeniy Riabenko, Milan Vojnovic
Carole-Jean Wu, Ramya Raghavendra, Udit Gupta, Bilge Acun, Newsha Ardalani, Kiwan Maeng, Gloria Chang, Fiona Aga Behram, James Huang, Charles Bai, Michael Gschwind, Anurag Gupta, Myle Ott, Anastasia Melnikov, Salvatore Candido, David Brooks, Geeta Chauhan, Benjamin Lee, Hsien-Hsin S. Lee, Bugra Akyildiz, Max Balandat, Joe Spisak, Ravi Jain, Mike Rabbat, Kim Hazelwood
Liqi Yan, Qifan Wang, Yiming Cu, Fuli Feng, Xiaojun Quan, Xiangyu Zhang, Dongfang Liu