More Like this

1. On Identifying Hashtags in Disaster Twitter Data

   https://doi.org/10.1609/aaai.v34i01.5387  

   Ray Chowdhury, Jishnu; Caragea, Cornelia; Caragea, Doina (June 2020, Proceedings of the AAAI Conference on Artificial Intelligence)

   null (Ed.)

   Tweet hashtags have the potential to improve the search for information during disaster events. However, there is a large number of disaster-related tweets that do not have any user-provided hashtags. Moreover, only a small number of tweets that contain actionable hashtags are useful for disaster response. To facilitate progress on automatic identification (or extraction) of disaster hashtags for Twitter data, we construct a unique dataset of disaster-related tweets annotated with hashtags useful for filtering actionable information. Using this dataset, we further investigate Long Short-Term Memory-based models within a Multi-Task Learning framework. The best performing model achieves an F1-score as high as $92.22%$. The dataset, code, and other resources are available on Github.1 

   more » « less
2. 2016-2021 State Hazard Mitigation Plans (SHMPs) and Social Vulnerability Qualitative Data:Subtitle

   https://doi.org/10.17603/ds2-g8tz-qg93  

   Peek, Lori; Painter, Mary Angelica; Villarreal, Melissa; Singh, Chandrika (June 2024, Designsafe-CI)

   U.S. state, territorial, and tribal government officials develop State Hazard Mitigation Plans (SHMPs) to assist in reducing the risk of disaster impacts on people, physical infrastructure, and the natural environment. The Federal Emergency Management Agency (FEMA) approves SHMPs every five years as a requirement to be eligible to receive funding for FEMA disaster relief grants and disaster mitigation projects. As of April 2023, updated FEMA policy guidance for SHMPs is in effect that calls for greater community engagement in the planning process and stipulates that plans consider equity and climate change. In response to these changes, this project takes the position that more robust conceptualizations of socially vulnerable populations and inclusive use of social vulnerability data can help states in the development of multi-hazard risk assessments. Social vulnerability emerges from systemic inequities, resulting in populations facing disproportionate impacts in disasters. It is a helpful framework for identifying underserved and marginalized populations. Given the crucial importance of considering social vulnerability in mitigation planning, our research team developed two novel datasets with descriptive data of the populations, definitions, and different measures of social vulnerability included in SHMPs for all 50 states and 5 inhabited U.S. territories. Specifically, this project includes two datasets: (1) a quantitative dataset where mentions of socially vulnerable populations and concepts are marked with a binary indicator of inclusion or exclusion in the State Hazard Mitigation Plan and (2) a qualitative dataset that contains quotes and locations of populations and concepts throughout each SHMP. The corresponding mission for each dataset includes: (1) the State Hazard Mitigation Plan dataset; (2) a data dictionary with description of each variable output; (3) variable definitions for the population groups included in State Hazard Mitigation Plans; and (4) a READ ME file with important information. These datasets and associated materials can help State Hazard Mitigation Officers and their technical partners identify gaps in addressing social vulnerability as they update the SHMPs for the areas they serve. These resources are available to researchers, practitioners, policy makers, and others who are interested in addressing social vulnerability in hazard mitigation planning. 

   more » « less
3. How Valid Are Social Vulnerability Models?

   Rufat, S.; Tate, E.; Emrich, C.; and Antolini, F. (March 2019, Annals of the American Association of Geographers)

   Social vulnerability models are becoming increasingly important for hazard mitigation and recovery planning,but it remains unclear how well they explain disaster outcomes. Most studies using indicators and indexes employ them to either describe vulnerability patterns or compare newly devised measures to existing ones. The focus of this article is construct validation, in which we investigate the empirical validity of a range of models of social vulnerability using outcomes from Hurricane Sandy. Using spatial regression, relative measures of assistance applicants, affected renters, housing damage, and property loss were regressed on four social vulnerability models and their constituent pillars while controlling for flood exposure. The indexes best explained housing assistance applicants, whereas they poorly explained property loss. At the pillar level,themes related to access and functional needs, age, transportation, and housing were the most explanatory.Overall, social vulnerability models with weighted and profile configurations demonstrated higher construct validity than the prevailing social vulnerability indexes. The findings highlight the need to expand the number and breadth of empirical validation studies to better understand relationships among social vulnerability models and disaster outcomes. 

   more » « less
4. Multimodal Semi-supervised Learning for Disaster Tweet Classification

   Iustin Sirbu, Tiberiu Sosea (October 2022, The 29th International Conference on Computational Linguistics (COLING 2022))

   During natural disasters, people often use social media platforms, such as Twitter, to post information about casualties and damage produced by disasters. This information can help relief authorities gain situational awareness in nearly real time, and enable them to quickly distribute resources where most needed. However, annotating data for this purpose can be burdensome, subjective and expensive. In this paper, we investigate how to leverage the copious amounts of unlabeled data generated on social media by disaster eyewitnesses and affected individuals during disaster events. To this end, we propose a semi-supervised learning approach to improve the performance of neural models on several multimodal disaster tweet classification tasks. Our approach shows significant improvements, obtaining up to 7.7% improvements in F-1 in low-data regimes and 1.9% when using the entire training data. We make our code and data publicly available at https://github.com/iustinsirbu13/multimodal-ssl-for-disaster-tweet-classification. 

   more » « less
5. Social Vulnerability across the Great Lakes Basin: A County-Level Comparative and Spatial Analysis

   https://doi.org/10.3390/su13137274  

   Fergen, Joshua T.; Bergstrom, Ryan D. (July 2021, Sustainability)

   Social vulnerability refers to how social positions affect the ability to access resources during a disaster or disturbance, but there is limited empirical examination of its spatial patterns in the Great Lakes Basin (GLB) region of North America. In this study, we map four themes of social vulnerability for the GLB by using the Center for Disease Control’s Social Vulnerability Index (CDC SVI) for every county in the basin and compare mean scores for each sub-basin to assess inter-basin differences. Additionally, we map LISA results to identify clusters of high and low social vulnerability along with the outliers across the region. Results show the spatial patterns depend on the social vulnerability theme selected, with some overlapping clusters of high vulnerability existing in Northern and Central Michigan, and clusters of low vulnerability in Eastern Wisconsin along with outliers across the basins. Differences in these patterns also indicate the existence of an urban–rural dimension to the variance in social vulnerabilities measured in this study. Understanding regional patterns of social vulnerability help identify the most vulnerable people, and this paper presents a framework for policymakers and researchers to address the unique social vulnerabilities across heterogeneous regions. 

   more » « less