Abstract: We present Tweet4act, a system to detect and classify crisis-related messages communicated over a microblogging platform. Our system relies on extracting content features from each message. These features and the use of an incident-specific dictionary allow us to determine the period type of an incident that each message belongs to. The period types are: Pre-incident (messages talking about prevention, mitigation, and preparedness), during-incident (messages sent while the incident is taking place), and post-incident (messages related to the response, recovery, and reconstruction). We show that our detection method can effectively identify incident-related messages with high precision and recall, and that our incident-period classification method outperforms standard machine learning classification methods.

Abstract: As the third hurricane the U.S. experienced in 2017, Hurricane María generated impacts that resulted in both short term and long term suffering in Puerto Rico. In this study, we aim to quantify the vulnerability of Puerto Ricans by taking region and society specific characteristics of the island into account. To do this, we follow Cutter et al.'s social vulnerability calculation, which is an inductive approach that aims to represent a society based on its characteristics. We adapted the Social Vulnerability Index (SoVI) for Puerto Rico by using data obtained from the U.S. Census Bureau. We analyzed the newly calculated SoVI for Puerto Rico and compared it with the existing deductive approach developed by the Center for Disease Control (CDC). Our findings show that the new index is able to capture some characteristics that the existing vulnerability index is unable to do.

Abstract: The Emergency Response domain is a highly challenging domain, requiring the active collaboration of several experts and authorities on the one hand and large-scale data analysis on the other. This poses significant challenges in sharing and analysing highly dynamic data describing highly evolving situations. This paper provides a brief summary for the first workshop in the SAFE workshop series. The workshop is aimed at bringing together analysts, practitioners, researchers and enthusiasts and provides a discussion ground for practical problems, solutions and projects that exploit Semantic Web, Linked Data analytics for Emergency Response. Following a round of thorough reviews, four papers are accepted and a keynote will complement the paper presentations along with a few discussion sessions.

Abstract: Like Crisis Management (CM) itself, Visual Analytics (VA) is a multi-disciplinary research area and is potentially useful to analyze and understand the huge amount of multidimensional data produced in a crisis. Our work investigates how researchers and practitioners are using VA in decision-making in CM. This paper firstly reports on a systematic mapping study to analyze the available information visualization tools and their applications in CM. To complement this information, we report on questionnaires and ethnographic studies applied during the large events held in Brazil in recent years. Then, we analyzed existing tools for visualizing crisis information. Lastly, we analyzed the data gathered from interviews with six professional crisis managers. The compiled results show that the full potential of VA is not being applied in the state-of-the-art and state-of-the-practice. We consider that further researches in the application of VA is required to improve decision-making processes in crisis management.

Abstract: Crisis Management (CM) refers to the ability to deal with crisis tasks in different phases and iterations. People working in a crisis are generally under pressure to make the right decision at the right time. They must process large amounts of data and assimilate the received information in an intuitive way. Visual Analytics (VA) is potentially useful to analyze and understand the huge amount of data in several areas including in a crisis. We propose a model based on VA to support decision-making in CM. The aim of the model is to help visualization designers to create effective VA interfaces, to help crisis managers to make quick and assertive decisions with them. In previous studies, we carried out a survey protocol with a multi-method approach to collect data on crisis related decision-making and analyze all these data qualitatively with formal techniques during the large events held in Brazil in recent years. In this work, we used our previous findings to develop the proposed model. We validated it using the focus group technique. With the new findings, we identified relevant insights on the use of VA for crisis management. We hope that, with these continuous cycles of validation and improvement, the agencies that manage crises might use our model as a reference for building more effective IT decision-making infrastructures based on VA.

Abstract: Whether it’s an infectious disease, a natural disaster, a human-made disaster, or a loss in utilities and resources, state and local leaders need visibility into the real-time resources of the entire healthcare continuum from labs, hospitals, long-term care settings, and shelters. By connecting public health and healthcare systems, information, and resources, leaders can be more agile and predictive in where to deploy limited resources before and during an emerging situation. The panelists will discuss how technology and data analytics can be utilized in real-time to resource decisions, bi-directional communication, transparency to stakeholders, and policy development. They will also explore the public health and healthcare continuum for mutual strategy, predictive modeling and reduction of excess loss of life. The panel will consist of a short introduction by each panelist followed by a facilitated discussion, and questions from the audience.

Abstract: Human-centered sensing (HCS) is an emerging research field that leverages mobile devices carried by people to collect useful information in support of myriad analytic activities. In this paper, we explore ways in which HCS can be applied to support a variety of analytic campaigns in the context of crisis response and management (CRM). We first summarize the concept of HCS and then investigate the potential advantages of complementing traditional sensing platforms and analytic tasks with an HCS system. By recognizing the potentials of HCS, we offer a scheme for classifying HCS systems and envision three application scenarios of HCS in CRM as well as a general architecture of HCS systems.

Abstract: The use of online Social Media is increasingly popular amongst emergency services to support Situational Awareness (i.e. accurate, complete and real-time information about an event). Whilst many software solutions have been developed to monitor and analyse Social Media, little attention has been paid on how to visually design for Situational Awareness for this large-scale data space. We describe an approach where levels of SA have been matched to corresponding visual design recommendations using participatory design techniques with Emergency Responders in the UK. We conclude by presenting visualisation prototypes developed to satisfy the design recommendations, and how they contribute to Emergency Responders' Situational Awareness in an example scenario. We end by highlighting research issues that emerged during the initial evaluation.

Abstract: The past decade has seen a significant increase in the use of social media for disaster management. This is due especially to the widespread usage of mobile devices and also to the different data types and data formats that social media supports. In recent years, research in the area of social media visual analytics has also gained interest in the scientific community. Research in this area however, lacks a comprehensive overview on social media visual analytics for disaster management. Hence, this paper presents a synthesis of extant research on social media visual analytic and visualization toolkits for disaster management. We survey available literature on these tools with the goal to outline the major characteristics and features, and to examine the extent to which they cover the full cycle of disaster management. Our main purpose is to provide a foundation based on the current literature that can help to shape future research directions to enhance social media visual analytic tools for full cycle disaster management.

Abstract: During the COVID-19 health crisis, local public offcials expend considerable energy encouraging citizens to comply with prevention measures in order to reduce the spread of infection. During the pandemic, mask-wearing has been accepted among health offcials as a simple preventative measure; however, some local areas have been more likely to comply than others. This paper explores methods to better understand local attitudes towards mask-wearing as a tool for public health offcials' situational awareness when preparing public messaging campaigns. This exploration compares three methods to explore local attitudes: sentiment analysis, n-grams, and hashtags. We also explore hashtag co-occurrence networks as a starting point to begin the filtering process. The results show that while sentiment analysis is quick and easy to employ, the results oer little insight into specific local attitudes towards mask-wearing, while examining hashtags and hashtag co-occurrence networks may be used a tool for a more robust understanding of local areas when attempting to gain situational awareness.

Abstract: The adoption of Artificial Intelligence (AI) technologies across various real-world applications for human performance augmentation demonstrates an unprecedented opportunity for emergency management. However, the current exploration of AI technologies such as computer vision and natural language processing is highly focused on emergency response and less investigated for the preparedness and mitigation phases. The training exercises for emergency services are critical to preparing responders to perform effectively in the real-world, providing a venue to leverage AI technologies. In this paper, we demonstrate an application of AI to address the challenges in augmenting the performance of instructors or trainers in such training exercises in real-time, with the explicit aim of reducing cognitive overload in extracting relevant knowledge from the voluminous multimodal data including video recordings and IoT sensor streams. We present an AI-infused system design for multimodal stream analytics and lessons from its use during a regional training exercise for active violence events.

Abstract: The recent rise of Social Media services has created huge streams of information which can be very valuable in a variety of scenarios. One specific scenario that has received interest is how Social Media analytics can be beneficial in crisis situations. In this paper, we describe our vision for a Social Media-ready command and control center. As motivation for our work, we present a short analysis of tweets issued in NYC during Hurricane Sandy in late October 2012 and we give an overview of the architecture of our event detection subsystem.

Abstract: The complex and changeful system of coal mine increases the difficulty and importance of its decision-making. Individual decisions sometimes can not bring satisfactory outcomes since the decision need broad knowledge and experience which is not in single field but related to many domains of economics, sociology, logic, etc. To improve the validity and objectivity of decision-making, the group decision-making method is feasible and necessary since it can collect more intelligence to choose and judge together. This paper synthetically analyzes the content and characteristic of decision-making in the field of coal mine safety. A methodology for group decision-making using analytic hierarchy process (AHP) with cluster analysis is proposed accordingly. Then a case study using the method indicates that it is effective and helpful to improve the level of decision-making in the field of coal mine safety management in China.