An official website of the United States government
Social media platforms are playing increasingly critical roles in disaster response and rescue operations. During emergencies, users can post rescue requests along with their addresses on social media, while volunteers can search for those messages and send help. However, efficiently leveraging social media in rescue operations remains challenging because of the lack of tools to identify rescue request messages on social media automatically and rapidly. Analyzing social media data, such as Twitter data, relies heavily on Natural Language Processing (NLP) algorithms to extract information from texts. The introduction of bidirectional transformers models, such as the Bidirectional Encoder Representations from Transformers (BERT) model, has significantly outperformed previous NLP models in numerous text analysis tasks, providing new opportunities to precisely understand and classify social media data for diverse applications. This study developed and compared ten VictimFinder models for identifying rescue request tweets, three based on milestone NLP algorithms and seven BERT-based. A total of 3191 manually labeled disaster-related tweets posted during 2017 Hurricane Harvey were used as the training and testing datasets. We evaluated the performance of each model by classification accuracy, computation cost, and model stability. Experiment results show that all BERT-based models have significantly increased the accuracy of categorizing rescue-related tweets. The best model for identifying rescue request tweets is a customized BERT-based model with a Convolutional Neural Network (CNN) classifier. Its F1-score is 0.919, which outperforms the baseline model by 10.6%. The developed models can promote social media use for rescue operations in future disaster events.
more »
« less
Social-Geographical Patterns of Rescue Requests During Hurricane Harvey
We analyze a public dataset of rescue requests for the Houston Metropolitan Area during Hurricane Harvey (2017) from the Red Cross. This dataset contains information including the location, gender, and emergency description in each requester’s report. We reveal the spatial distribution of the rescue requests and its relationship with indicators of the social, physical, and built environment. We show that the rescue request rates are significantly higher in regions with higher percentages of children, male population, population in poverty, or people with limited English, in addition to regions with higher inundation rate or worse traffic condition during Hurricane Harvey. The rescue request rate is found to be statistically uncorrelated with the percentage of flood hazard zone designated by the Federal Emergency Management Agency (FEMA).
more »
« less
- Award ID(s):
- 1652448
- PAR ID:
- 10481171
- Publisher / Repository:
- Findings
- Date Published:
- Journal Name:
- Findings
- ISSN:
- 2652-8800
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Hurricane Harvey in 2017 marked an important transition where many disaster victims used social media rather than the overloaded 911 system to seek rescue. This article presents a machine-learning-based detector of rescue requests from Harvey-related Twitter messages, which differentiates itself from existing ones by accounting for the potential impacts of ZIP codes on both the preparation of training samples and the performance of different machine learning models. We investigate how the outcomes of our ZIP code filtering differ from those of a recent, comparable study in terms of generating training data for machine learning models. Following this, experiments are conducted to test how the existence of ZIP codes would affect the performance of machine learning models by simulating different percentages of ZIP-code-tagged positive samples. The findings show that (1) all machine learning classifiers except K-nearest neighbors and Naïve Bayes achieve state-of-the-art performance in detecting rescue requests from social media; (2) using ZIP code filtering could increase the effectiveness of gathering rescue requests for training machine learning models; (3) machine learning models are better able to identify rescue requests that are associated with ZIP codes. We thereby encourage every rescue-seeking victim to include ZIP codes when posting messages on social media. This study is a useful addition to the literature and can be helpful for first responders to rescue disaster victims more efficiently.more » « less
-
Abstract AI fairness is tasked with evaluating and mitigating bias in algorithms that may discriminate towards protected groups. This paper examines if bias exists in AI algorithms used in disaster management and in what manner. We consider the 2017 Hurricane Harvey when flood victims in Houston resorted to social media to request for rescue. We evaluate a Random Forest regression model trained to predict Twitter rescue request rates from social-environmental data using three fairness criteria (independence, separation, and sufficiency). The Social Vulnerability Index (SVI), its four sub-indices, and four variables representing digital divide were considered sensitive attributes. The Random Forest regression model extracted seven significant predictors of rescue request rates, and from high to low importance they were percent of renter occupied housing units, percent of roads in flood zone, percent of flood zone area, percent of wetland cover, percent of herbaceous, forested and shrub cover, mean elevation, and percent of households with no computer or device. Partial Dependence plots of rescue request rates against each of the seven predictors show the non-linear nature of their relationships. Results of the fairness evaluation of the Random Forest model using the three criteria show no obvious biases for the nine sensitive attributes, except that a minor imperfect sufficiency was found with the SVI Housing and Transportation sub-index. Future AI modeling in disaster research could apply the same methodology used in this paper to evaluate fairness and help reduce unfair resource allocation and other social and geographical disparities.more » « less
-
Accurate flood forecasting and efficient emergency response operations are vital, especially in the case of urban flash floods. The dense distribution of power lines in urban areas significantly impacts search and rescue operations during extreme flood events. However, no existing emergency response frameworks have incorporated the impacts of overhead power lines on lifeboat rescue operations. This study aims to determine the necessity and feasibility of incorporating overhead power line information into an emergency response framework using Manville, New Jersey during Hurricane Ida as a test bed. We propose an integrated framework, which includes a building-scale flood model, urban point cloud data, a human vulnerability model, and network analysis, to simulate rescue operation feasibility during Hurricane Ida. Results reveal that during the most severe point of the flood event, 46% of impacted buildings became nonrescuable due to complete isolation from the road network, and a significant 67.7% of the municipality’s areas that became dangerous for pedestrians also became inaccessible to rescue boats due to overhead power line obstruction. Additionally, we identify a continuous 10-hour period during which an average of 43.4% of the 991 impacted buildings faced complete isolation. For these structures, early evacuation emerges as the sole means to prevent isolation. This research highlights the pressing need to consider overhead power lines in emergency response planning to ensure more effective and targeted flood resilience measures for urban areas facing increasingly frequent extreme precipitation events.more » « less
-
During flash flooding, quick and effective rescue operations are crucial to minimizing harm to vulnerable communities. While much research focused on emergency response and evacuation, few studies address how overhead powerline obstructions impact rescue operations. Additionally, existing research on vulnerable communities often emphasizes long-term flood mitigation and recovery but less so on immediate responses. To ensure rapid and equitable flood rescue operations, this study derives an integrated metric to quantify rescue demands that incorporate rescue efficiency, community flood severity, and social vulnerability. In detail, rescue efficiency is calculated by analyzing a network that captures the geospatial interdependencies between the residential buildings' road networks and overhead power lines; community flood severity is quantified as the percentage of building damage resulting from flood impacts; and social vulnerability is an integrated indication of key household composition factors (e.g., elders, single parents, and minorities). Based on this metric, a systematic step is designed to suggest the sequence of rescue operations and the strategies for distributing rescue lifeboats at emergency facilities. The applicability and feasibility of the proposed approach were demonstrated using lifeboat rescue operations in Manville, New Jersey, during Hurricane Ida. This study calculates dynamic changes in rescue loads of all emergency facilities and then finds the optimal strategies for distributing lifeboats. The results highlight the significant impact of overhead power line obstructions on the optimal rescue lifeboat distribution. Additionally, the results suggest prioritizing emergency evacuation for socially vulnerable households in Manville township. Practically, the generated rescue sequence and rescue lifeboat distribution are expected to help emergency response agencies perform effective and rapid rescue operations.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.32866/001c.37903
