Coupling sentiment and human mobility in natural disasters: a Twitter-based study of the 2014 South Napa Earthquake

Understanding population dynamics during natural disasters is important to build urban resilience in preparation for extreme events. Social media has emerged as an important source for disaster managers to identify dynamic polarity of sentiments over the course of disasters, to understand human mobility patterns, and to enhance decision making and disaster recovery efforts. Although there is a growing body of literature on sentiment and human mobility in disaster contexts, the spatiotemporal characteristics of sentiment and the relationship between sentiment and mobility over time have not been investigated in detail. This study therefore addresses this research gap and proposes a new lens to evaluate population dynamics during disasters by coupling sentiment and mobility. We collected 3.74 million geotagged tweets over 8 weeks to examine individuals’ sentiment and mobility before, during and after the M6.0 South Napa, California Earthquake in 2014. Our research results reveal that the average sentiment level decreases with the increasing intensity of the earthquake. We found that similar levels of sentiment tended to cluster in geographical space, and this spatial autocorrelation was significant over areas of different earthquake intensities. Moreover, we investigated the relationship between temporal dynamics of sentiment and mobility. We examined the trend and seasonality of the time series and found cointegration between the series. We included effects of the earthquake and built a segmented regression model to describe the time series finding that day-to-day changes in sentiment can either lead or lag daily changed mobility patterns. This study contributes a new lens to assess the dynamic process of disaster resilience unfolding over large spatial scales.

[1]  Jeannie A. Stamberger,et al.  Crowd sentiment detection during disasters and crises , 2012, ISCRAM.

[2]  David R. Godschalk,et al.  Urban Hazard Mitigation: Creating Resilient Cities , 2003 .

[3]  Guang Yu,et al.  A Weibo-based approach to disaster informatics: incidents monitor in post-disaster situation via Weibo text negative sentiment analysis , 2016, Natural Hazards.

[4]  Petter Holme,et al.  Predictability of population displacement after the 2010 Haiti earthquake , 2012, Proceedings of the National Academy of Sciences.

[5]  Nigel Collier,et al.  Twitter Emotion Analysis in Earthquake Situations , 2013, Int. J. Comput. Linguistics Appl..

[6]  Xiangyang Guan,et al.  Using social media data to understand and assess disasters , 2014, Natural Hazards.

[7]  P. Moran.  Notes on continuous stochastic phenomena. , 1950, Biometrika.

[8]  H. Hong,et al.  Using remote sensing information to estimate snow hazard and extreme snow load in China , 2017, Natural Hazards.

[9]  Christopher M. Danforth,et al.  The Geography of Happiness: Connecting Twitter Sentiment and Expression, Demographics, and Objective Characteristics of Place , 2013, PloS one.

[10]  Qi Wang,et al.  Process Map for Urban-Human Mobility and Civil Infrastructure Data Collection Using Geosocial Networking Platforms , 2016, J. Comput. Civ. Eng..

[11]  Christopher M. Danforth,et al.  Happiness and the Patterns of Life: A Study of Geolocated Tweets , 2013, Scientific Reports.

[12]  Marta C. González,et al.  Understanding individual human mobility patterns , 2008, Nature.

[13]  Fabrício Benevenuto,et al.  A Benchmark Comparison of State-of-the-Practice Sentiment Analysis Methods , 2015, ArXiv.

[14]  Qi Wang,et al.  Quantifying Human Mobility Perturbation and Resilience in Hurricane Sandy , 2014, PloS one.

[15]  P. Cowpertwait,et al.  Introductory Time Series with R , 2009 .

[16]  Cornelia Caragea,et al.  Sentiment analysis during Hurricane Sandy in emergency response , 2017 .

[17]  Yu-Ru Lin,et al.  Assessing Sentiment Segregation in Urban Communities , 2014, SocialCom '14.

[18]  C. Granger.  Some properties of time series data and their use in econometric model specification , 1981 .

[19]  D. Dickey,et al.  Testing for unit roots in autoregressive-moving average models of unknown order , 1984 .

[20]  Christopher M. Danforth,et al.  Climate Change Sentiment on Twitter: An Unsolicited Public Opinion Poll , 2015, PloS one.

[21]  Witold Pedrycz,et al.  Sentiment Analysis and Ontology Engineering - An Environment of Computational Intelligence , 2016, Sentiment Analysis and Ontology Engineering.

[22]  Zhenghong Tang,et al.  Examining the role of social media in California’s drought risk management in 2014 , 2015, Natural Hazards.

[23]  J. Zhuang,et al.  Crisis information distribution on Twitter: a content analysis of tweets during Hurricane Sandy , 2017, Natural Hazards.

[24]  Maximilian Walther,et al.  Geo-spatial Event Detection in the Twitter Stream , 2013, ECIR.

[25]  Ghazaleh Beigi,et al.  An Overview of Sentiment Analysis in Social Media and Its Applications in Disaster Relief , 2016, Sentiment Analysis and Ontology Engineering.

[26]  P. Phillips,et al.  Asymptotic Properties of Residual Based Tests for Cointegration , 1990 .

[27]  J. Taylor,et al.  Aggregated responses of human mobility to severe winter storms: An empirical study , 2017, PloS one.

[28]  J. Fowler,et al.  Rapid assessment of disaster damage using social media activity , 2016, Science Advances.

[29]  Finn Årup Nielsen,et al.  A New ANEW: Evaluation of a Word List for Sentiment Analysis in Microblogs , 2011, #MSM.

[30]  A K Wagner,et al.  Segmented regression analysis of interrupted time series studies in medication use research , 2002, Journal of clinical pharmacy and therapeutics.

[31]  John E. Taylor,et al.  Patterns and Limitations of Urban Human Mobility Resilience under the Influence of Multiple Types of Natural Disaster , 2016, PloS one.

[32]  Chen Chen,et al.  Opinion: Big data has big potential for applications to climate change adaptation , 2016, Proceedings of the National Academy of Sciences.

[33]  R. Geary,et al.  The Contiguity Ratio and Statistical Mapping , 1954 .

[34]  Jacob Eisenstein,et al.  Confounds and Consequences in Geotagged Twitter Data , 2015, EMNLP.

[35]  D. Ballas,et al.  The Geography of Happiness , 2013 .

[36]  Ming-Hsiang Tsou,et al.  Spatial, temporal, and content analysis of Twitter for wildfire hazards , 2016, Natural Hazards.