|
Antone Evans Jr., Yingyuan Yang, & Sunshin Lee. (2021). Towards Predicting COVID-19 Trends: Feature Engineering on Social Media Responses. In Anouck Adrot, Rob Grace, Kathleen Moore, & Christopher W. Zobel (Eds.), ISCRAM 2021 Conference Proceedings – 18th International Conference on Information Systems for Crisis Response and Management (pp. 792–807). Blacksburg, VA (USA): Virginia Tech.
Abstract: During the course of this pandemic, the use of social media and virtual networks has been at an all-time high. Individuals have used social media to express their thoughts on matters related to this pandemic. It is difficult to predict current trends based on historic case data because trends are more connected to social activities which can lead to the spread of coronavirus. So, it's important for us to derive meaningful information from social media as it is widely used. Therefore, we grouped tweets by common keywords, found correlations between keywords and daily COVID-19 statistics and built predictive modeling. The features correlation analysis was very effective, so trends were predicted very well. A RMSE score of 0.0425504, MAE of 0.03295105 and RSQ of 0.5237014 in relation to daily deaths. In addition, we found a RMSE score of 0.07346836, MAE of 0.0491152 and RSQ 0.374529 in relation to daily cases.
|
|
|
Daniel Link, Bernd Hellingrath, & Jie Ling. (2016). A Human-is-the-Loop Approach for Semi-Automated Content Moderation. In A. Tapia, P. Antunes, V.A. Bañuls, K. Moore, & J. Porto (Eds.), ISCRAM 2016 Conference Proceedings ? 13th International Conference on Information Systems for Crisis Response and Management. Rio de Janeiro, Brasil: Federal University of Rio de Janeiro.
Abstract: Online social media has been recognized as a valuable information source for disaster management whose volume, velocity and variety exceed manual processing capacity. Current machine learning systems that support the processing of such data generally follow a human-in-the-loop approach, which has several inherent limitations. This work applies the human-is-the-loop concept from visual analytics to semi-automate a manual content moderation workflow, wherein human moderators take the dominant role. The workflow is instantiated with a supervised machine learning system that supports moderators with suggestions regarding the relevance and categorization of content. The instantiated workflow has been evaluated using in-depth interviews with practitioners and serious games. which suggest that it offers good compatibility with work practices in humanitarian assessment as well as improved moderation quality and higher flexibility than common approaches.
|
|
|
Federico Angaramo, & Claudio Rossi. (2018). Online clustering and classification for real-time event detection in Twitter. In Kees Boersma, & Brian Tomaszeski (Eds.), ISCRAM 2018 Conference Proceedings – 15th International Conference on Information Systems for Crisis Response and Management (pp. 1098–1107). Rochester, NY (USA): Rochester Institute of Technology.
Abstract: Event detection from social media is a challenging task due to the volume, the velocity and the variety of user-generated data requiring real-time processing. Despite recent works on this subject, a generalized and scalable approach that could be applied across languages and topics has not been consolidated, yet. In this paper, we propose a methodology for real-time event detection from Twitter data that allows users to select a topic of interest by defining a simple set of keywords and a matching rule. We implement the proposed methodology and evaluate it with real data to detect different types of events.
|
|
|
Florian Vandecasteele, Krishna Kumar, Kenzo Milleville, & Steven Verstockt. (2019). Video Summarization And Video Highlight Selection Tools To Facilitate Fire Incident Management. In Z. Franco, J. J. González, & J. H. Canós (Eds.), Proceedings of the 16th International Conference on Information Systems for Crisis Response And Management. Valencia, Spain: Iscram.
Abstract: This paper reports on the added value of combining different types of sensor data and geographic information for fire incident management. A survey was launched within the Belgian fire community to explore the need of added value and the use of new types of sensor data during a fire incident. This evaluation revealed that people are visually-oriented and that video footages and images are of great value to gain insights in a particular problem. However, due to the limited available time (i.e., fast decisions need to be taken) and the large amount of cameras it is not feasible to analyze all video footages sequentially. To solve this problem we propose a video summarization mechanism and a video highlight selection tool based on the automatic generated image and video tags.
|
|
|
Hristo Tanev, Vanni Zavarella, & Josef Steinberger. (2017). Monitoring disaster impact: detecting micro-events and eyewitness reports in mainstream and social media. In eds Aurélie Montarnal Matthieu Lauras Chihab Hanachi F. B. Tina Comes (Ed.), Proceedings of the 14th International Conference on Information Systems for Crisis Response And Management (pp. 592–602). Albi, France: Iscram.
Abstract: This paper approaches the problem of monitoring the impact of the disasters by mining web sources for the events, caused by these disasters. We refer to these disaster effects as “micro-events”. Micro-events typically following a large disaster include casualties, damage on infrastructures, vehicles, services and resource supply, as well as relief operations. We present natural language grammar learning algorithms which form the basis for building micro-event detection systems from data, with no or minor human intervention, and we show how they can be applied to mainstream news and social media for monitoring disaster impact. We also experimented with applying statistical classifiers to distill, from social media situational updates on disasters, eyewitness reports from directly affected people. Finally, we describe a Twitter mining robot, which integrates some of these monitoring techniques and is intended to serve as a multilingual content hub for enhancing situational awareness.
|
|
|
Muhammad Imran, Shady Elbassuoni, Carlos Castillo, Fernando Díaz, & Patrick Meier. (2013). Extracting information nuggets from disaster- Related messages in social media. In J. Geldermann and T. Müller S. Fortier F. F. T. Comes (Ed.), ISCRAM 2013 Conference Proceedings – 10th International Conference on Information Systems for Crisis Response and Management (pp. 791–801). KIT; Baden-Baden: Karlsruher Institut fur Technologie.
Abstract: Microblogging sites such as Twitter can play a vital role in spreading information during “natural” or man-made disasters. But the volume and velocity of tweets posted during crises today tend to be extremely high, making it hard for disaster-affected communities and professional emergency responders to process the information in a timely manner. Furthermore, posts tend to vary highly in terms of their subjects and usefulness; from messages that are entirely off-topic or personal in nature, to messages containing critical information that augments situational awareness. Finding actionable information can accelerate disaster response and alleviate both property and human losses. In this paper, we describe automatic methods for extracting information from microblog posts. Specifically, we focus on extracting valuable “information nuggets”, brief, self-contained information items relevant to disaster response. Our methods leverage machine learning methods for classifying posts and information extraction. Our results, validated over one large disaster-related dataset, reveal that a careful design can yield an effective system, paving the way for more sophisticated data analysis and visualization systems.
|
|
|
Laura Szczyrba, Yang Zhang, Duygu Pamukcu, & Derya Ipek Eroglu. (2020). A Machine Learning Method to Quantify the Role of Vulnerability in Hurricane Damage. In Amanda Hughes, Fiona McNeill, & Christopher W. Zobel (Eds.), ISCRAM 2020 Conference Proceedings – 17th International Conference on Information Systems for Crisis Response and Management (pp. 179–187). Blacksburg, VA (USA): Virginia Tech.
Abstract: Accurate pre-disaster damage predictions and post-disaster damage assessments are challenging because of the complicated interrelationships between multiple damage drivers, including various natural hazards, as well as antecedent infrastructure quality and demographic characteristics. Ensemble decision trees, a family of machine learning algorithms, are well suited to quantify the role of social vulnerability in disaster impacts because they provide interpretable measures of variable importance for predictions. Our research explores the utility of an ensemble decision tree algorithm, Random Forest Regression, for quantifying the role of vulnerability with a case study of Hurricane Mar\'ia. The contributing predictive power of eight drivers of structural damage was calculated as the decrease in model mean squared error. A measure of social vulnerability was found to be the model's leading predictor of damage patterns. An additional algorithm, other methods of quantifying variable importance, and future work are discussed.
|
|
|
López-Catalán, B., & Bañuls, V. A. (2023). A Topic Modeling Approach for Extracting Key City Resilience Indicators. In Jaziar Radianti, Ioannis Dokas, Nicolas Lalone, & Deepak Khazanchi (Eds.), Proceedings of the 20th International ISCRAM Conference (pp. 944–952). Omaha, USA: University of Nebraska at Omaha.
Abstract: In the field of urban resilience, there is a great diversity of approaches to measuring the level of resilience in cities. This information is scattered among reports and academic articles. In this ongoing research paper, we explore the potential of Topic Modeling to analyze this information, in order to determine cluster indicators for a set of academic papers and resilience frameworks. These clusters are referred to as Key City Resilience Indicators (KCRI), which are used as reference to facilitate the measurement of urban resilience regardless of the context, including all the key dimensions required for cities to achieve resilience. Topic modeling outcomes can be used to generate indicators based on each topic or to automatically classify a new set of indicators in each of the established topics. These results can be applied to any resilience framework
|
|
|
Md Fitrat Hossain, Thomas Kissane, Priyanka Annapureddy, Wylie Frydrychowicz, Sheikh Iqbal Ahamed, Naveen Bansal, et al. (2020). Implementing Algorithmic Crisis Alerts in mHealth Systems for Veterans with PTSD. In Amanda Hughes, Fiona McNeill, & Christopher W. Zobel (Eds.), ISCRAM 2020 Conference Proceedings – 17th International Conference on Information Systems for Crisis Response and Management (pp. 122–133). Blacksburg, VA (USA): Virginia Tech.
Abstract: This paper seeks to establish a machine learning driven method by which a military veteran with Post-Traumatic Stress Disorder (PTSD) is classified as being in a crisis situation or not, based upon a given set of criteria. Optimizing alerting decision rules is critical to ensure that veterans at highest risk for mental health crisis rapidly receive additional attention. Subject matter experts in our team (a psychologist, a medical anthropologist, and an expert veteran), defined acute crisis, early warning signs and long-term crisis from this dataset. First, we used a decision tree to find an early time point when the peer mentors (who are also veterans) need to observe the behavior of veterans to make a decision about conducting an intervention. Three different machine learning algorithms were used to predict long term crisis using acute crisis and early warning signs within the determined time point.
|
|
|
Niels Netten, & Maarten Van Someren. (2006). Automated support for dynamic information distribution in incident management. In M. T. B. Van de Walle (Ed.), Proceedings of ISCRAM 2006 – 3rd International Conference on Information Systems for Crisis Response and Management (pp. 230–237). Newark, NJ: Royal Flemish Academy of Belgium.
Abstract: For all emergency response personnel involved in crisis situations it is essential to timely acquire all information critical to their task performance. However, in practice errors occur in the distribution of information between these collaborating actors leading to mistakes and subsequently more damage to the situation. In this paper we present a prototype system for dynamic information distribution able to support the information flow between collaborating crisis actors. The system has been evaluated by means of simulated experiments that use data from a real incident scenario. The results indicate that automated support by means of Machine Learning method works well. Especially, when actor work context features are included, then the performance on selecting and distributing relevant information is high. Furthermore, actors acquire relevant information much faster making group communication much more efficient.
|
|
|
Nilani Algiriyage, Rangana Sampath, Raj Prasanna, Kristin Stock, Emma Hudson-Doyle, & David Johnston. (2021). Identifying Disaster-related Tweets: A Large-Scale Detection Model Comparison. In Anouck Adrot, Rob Grace, Kathleen Moore, & Christopher W. Zobel (Eds.), ISCRAM 2021 Conference Proceedings – 18th International Conference on Information Systems for Crisis Response and Management (pp. 731–743). Blacksburg, VA (USA): Virginia Tech.
Abstract: Social media applications such as Twitter and Facebook are fast becoming a key instrument in gaining situational awareness (understanding the bigger picture of the situation) during disasters. This has provided multiple opportunities to gather relevant information in a timely manner to improve disaster response. In recent years, identifying crisis-related social media posts is analysed as an automatic task using machine learning (ML) or deep learning (DL) techniques. However, such supervised learning algorithms require labelled training data in the early hours of a crisis. Recently, multiple manually labelled disaster-related open-source twitter datasets have been released. In this work, we create a large dataset with 186,718 tweets by combining a number of such datasets and evaluate the performance of multiple ML and DL algorithms in classifying disaster-related tweets in three settings, namely ``in-disaster'', ``out-disaster'' and ``cross-disaster''. Our results show that the Bidirectional LSTM model with Word2Vec embeddings performs well for the tweet classification task in all three settings. We also make available the preprocessing steps and trained weights for future research.
|
|
|
Patrick Lieser, Alaa Alhamoud, Hosam Nima, Björn Richerzhagen, Sanja Huhle, Doreen Böhnstedt, et al. (2018). Situation Detection based on Activity Recognition in Disaster Scenarios. In Kees Boersma, & Brian Tomaszeski (Eds.), ISCRAM 2018 Conference Proceedings – 15th International Conference on Information Systems for Crisis Response and Management (pp. 737–753). Rochester, NY (USA): Rochester Institute of Technology.
Abstract: In disaster situations like earthquakes and hurricanes, people have difficulties accessing shelter and requesting help. Many smartphone applications provide behavioral advice or means to communicate during such situations. However, to what extent a person is affected by a disaster is often unclear, as these applications rely on the user's subjective assessment. Therefore, detecting a user's situation is key to provide more meaningful information in such applications and to allows first responders to better assess incoming messages. We propose a predictive model that recognizes four normal and ten disaster-related activities achieving an average f1-score of up to 90.1\%, solely based on sensor readings of the subject's mobile device. We conduct an extensive measurement-based evaluation to assess the impact of individual model parameters on the prediction accuracy. Our model is orientation-independent, position-independent, and subject-independent, making it an ideal foundation for future context-aware emergency applications.
|
|
|
Peter Serwylo, Paul Arbon, & Grace Rumantir. (2011). Predicting patient presentation rates at mass gatherings using machine learning. In E. Portela L. S. M.A. Santos (Ed.), 8th International Conference on Information Systems for Crisis Response and Management: From Early-Warning Systems to Preparedness and Training, ISCRAM 2011. Lisbon: Information Systems for Crisis Response and Management, ISCRAM.
Abstract: Mass gatherings have been defined as events where more than 1,000 people are present for a defined period of time. Such an event presents specific challenges with respect to medical care. First aid is provisioned on-site at most events in order to prevent undue strain on the local emergency services. In order to allocate enough resources to deal with the expected injuries, it is important to be able to accurately predict patient volumes. This study used machine learning techniques to identify which variables are the most important in predicting patient volumes at mass gatherings. Data from 201 mass gatherings across Australia was analysed, finding that event type is the most predictive variable, followed by the state or territory, heat index, humidity, whether it is bounded, and the time of day. Variables with little bearing on the outcome included the presence of alcohol, whether the event was indoors or outdoors, and whether it had one point of focus. The best predictive models produced acceptable predictions of the patient presentations 80% of the time, and this could be further improved using optimization techniques.
|
|
|
Shivam Sharma, & Cody Buntain. (2021). An Evaluation of Twitter Datasets from Non-Pandemic Crises Applied to Regional COVID-19 Contexts. In Anouck Adrot, Rob Grace, Kathleen Moore, & Christopher W. Zobel (Eds.), ISCRAM 2021 Conference Proceedings – 18th International Conference on Information Systems for Crisis Response and Management (pp. 808–815). Blacksburg, VA (USA): Virginia Tech.
Abstract: In 2020, we have witnessed an unprecedented crisis event, the COVID-19 pandemic. Various questions arise regarding the nature of this crisis data and the impacts it would have on the existing tools. In this paper, we aim to study whether we can include pandemic-type crisis events with general non-pandemic events and hypothesize that including labeled crisis data from a variety of non-pandemic events will improve classification performance over models trained solely on pandemic events. To test our hypothesis we study the model performance for different models by performing a cross validation test on pandemic only held-out sets for two different types of training sets, one containing only pandemic data and the other a combination of pandemic and non-pandemic crisis data, and comparing the results of the two. Our results approve our hypothesis and give evidence of some crucial information propagation upon inclusion of non-pandemic crisis data to pandemic data.
|
|
|
Vitaveska Lanfranchi. (2017). Machine Learning and Social Media in Crisis Management: Agility vs Ethics. In eds Aurélie Montarnal Matthieu Lauras Chihab Hanachi F. B. Tina Comes (Ed.), Proceedings of the 14th International Conference on Information Systems for Crisis Response And Management (pp. 256–265). Albi, France: Iscram.
Abstract: One of the most used sources of information for fast and flexible crisis information is social media or crowdsourced data, as the information is rapidly disseminated, can reach a large amount of target audience and covers a wide variety of topics. However, the agility that these new methodologies enable comes at a price: ethics and privacy. This paper presents an analysis of the ethical risks and implications of using automated system that learn from social media data to provide intelligence in crisis management. The paper presents a short overview on the use of social media data in crisis management to then highlight ethical implication of machine learning and social media data using an example scenario. In conclusion general mitigation strategies and specific implementation guidelines for the scenario under analysis are presented.
|
|
|
Yingjie Li, Seoyeon Park, Cornelia Caragea, Doina Caragea, & Andrea Tapia. (2019). Sympathy Detection in Disaster Twitter Data. In Z. Franco, J. J. González, & J. H. Canós (Eds.), Proceedings of the 16th International Conference on Information Systems for Crisis Response And Management. Valencia, Spain: Iscram.
Abstract: Nowadays, micro-blogging sites such as Twitter have become powerful tools for communicating with others in
various situations. Especially in disaster events, these sites can be the best platforms for seeking or providing social
support, of which informational support and emotional support are the most important types. Sympathy, a sub-type
of emotional support, is an expression of one?s compassion or sorrow for a difficult situation that another person
is facing. Providing sympathy to people affected by a disaster can help change people?s emotional states from
negative to positive emotions, and hence, help them feel better. Moreover, detecting sympathy contents in Twitter
can potentially be used for finding candidate donors since the emotion ?sympathy? is closely related to people who
may be willing to donate. Thus, in this paper, as a starting point, we focus on detecting sympathy-related tweets.
We address this task using Convolutional Neural Networks (CNNs) with refined word embeddings. Specifically, we
propose a refined word embedding technique in terms of various pre-trained word vector models and show great
performance of CNNs that use these refined embeddings in the sympathy tweet classification task. We also report
experimental results showing that the CNNs with the refined word embeddings outperform not only traditional
machine learning techniques, such as Naïve Bayes, Support Vector Machines and AdaBoost with conventional
feature sets as bags of words, but also Long Short-Term Memory Networks.
|
|