|
Changwon Son, Jukrin Moon, S. Camille Peres, & Farzan Sasangohar. (2018). An Episode as a Trace of Resilient Performance in Multi-Agency Incident Management Systems. In Kees Boersma, & Brian Tomaszeski (Eds.), ISCRAM 2018 Conference Proceedings – 15th International Conference on Information Systems for Crisis Response and Management (pp. 942–948). Rochester, NY (USA): Rochester Institute of Technology.
Abstract: In order to cope with increasing complexity of catastrophic disasters, resilience is considered an essential capability of an incident management system (IMS). As resilience is manifested during systems operation, a naturalistic observational study was conducted to understand how resilient performance dynamically takes place in this domain. The study results were presented using the concept of episodes, each of which uncovers a trace of such resilient performance following an information input called an inject. The episode analysis also facilitated the identification of complex and dynamic interactions among human and technological agents to satisfy work demands, representing work-as-done (WAD) in large-scale emergency response operations.
|
|
|
Chanthujan Chandrakumar, Raj Prasanna, Max Stephens, Marion Lara Tan, Caroline Holden, Amal Punchihewa, et al. (2023). Algorithms for Detecting P-Waves and Earthquake Magnitude Estimation: Initial Literature Review Findings. In V. L. Thomas J. Huggins (Ed.), Proceedings of the ISCRAM Asia Pacific Conference 2022 (pp. 138–155). Palmerston North, New Zealand: Massey Unversity.
Abstract: Earthquake Early Warning System (EEWS) plays a major role during an earthquake in alerting the public and authorities to take appropriate safety measures during an earthquake. Generally, EEWSs use three types of algorithms to generate alerts during an earthquake; namely: source-based, ground motion or wavefield-based and on-site-based approaches. However, source-based algorithms are commonly used in most of EEWSs worldwide. A source-based EEWS uses a particular time frame of the P-wave of an earthquake to estimate the source parameters such as magnitude and the location of that earthquake with the support of P-wave detection and earthquake magnitude and location estimation algorithms. As the initial step of a research project which aims to explore the best use of P-waves to generate earthquake alerts, this Work in Progress paper (WiPe) presents the initial partial findings from an ongoing literature review on exploring the algorithms used for P-wave detection and earthquake magnitude estimation.
|
|
|
Chanvi Kotak, Brian Tomaszewski, & Erik Golen. (2018). 3-1-1 Calls Hot Spot Analysis During Hurricane Harvey: Preliminary Results. In Kees Boersma, & Brian Tomaszeski (Eds.), ISCRAM 2018 Conference Proceedings – 15th International Conference on Information Systems for Crisis Response and Management (pp. 350–361). Rochester, NY (USA): Rochester Institute of Technology.
Abstract: Hurricane Harvey caused massive damage and necessitated the need for identification of areas under high risk. During Harvey, the city of Houston received more than 77000, 3-1-1 calls for assistance. Due to damage caused to the infrastructure, it became difficult to handle and respond to the crisis. Geographic Information Systems (GIS) is a vital technology to assist with real-time disaster monitoring. we investigated if a correlation could be found between 311 data calls made during the Hurricane Harvey and aerial images captured during the event, specifically to see if 311 data could be ground-truthed via hot spot analysis. Preliminary results indicate that visual representation of 3-1-1 call data can aid in analyzing the expected areas of high traffic of calls for assistance and plan an effective way to manage resources. Future work will involve more in-depth analysis of combined 3-1-1 call data with satellite imagery using image classification techniques.
|
|
|
Charles Bailly, & Carole Adam. (2017). An interactive simulation for testing communication strategies in bushfires. 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. 72–84). Albi, France: Iscram.
Abstract: Australia is frequently hit by bushfires. In 2009, the “Black Saturday” fires killed 173 people and burnt hectares of bush. As a result, a research commission was created to investigate, and concluded that several aspects could be improved, in particular better understanding of the population actual behaviour, and better communication with them. We argue that agent-based modelling and simulation is a great tool to test possible communication strategies, in order to deduce valuable insight for emergency managers before new fires happen. In this paper, we extend an existing agent-based model of the population behaviour in bushfires. Concretely, we added a communication model based in social sciences, and user interactivity with the model. We present the results of first experiments with dierent communication strategies, providing valuable insight for better communication with the population during such events. This model is still preliminary and will eventually be turned into a serious game.
|
|
|
Chauhan, A. (2023). Humor-Based COVID-19 Twitter Accounts. In Jaziar Radianti, Ioannis Dokas, Nicolas Lalone, & Deepak Khazanchi (Eds.), Proceedings of the 20th International ISCRAM Conference (pp. 417–427). Omaha, USA: University of Nebraska at Omaha.
Abstract: Crisis Named Resources (or CNRs) are social media pages and accounts named after a crisis event. Using the COVID-19 Pandemic as a case study, we identified and examined the role of CNRs that shared humor on Twitter. Our analyses showed that humor-based CNRs shared virus-related rumors, stigma, safety measures, opinions, sarcasm, and news updates. These resources also shared the overall anger and frustration over the year 2020. We conclude by discussing the critical role of humor based CNRs in crisis response.
|
|
|
Cheng Wang, Benjamin Bowes, Arash Tavakoli, Stephen Adams, Jonathan Goodall, & Peter Beling. (2020). Smart Stormwater Control Systems: A Reinforcement Learning Approach. 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. 2–13). Blacksburg, VA (USA): Virginia Tech.
Abstract: Flooding poses a significant and growing risk for many urban areas. Stormwater systems are typically used to control flooding, but are traditionally passive (i.e. have no controllable components). However, if stormwater systems are retrofitted with valves and pumps, policies for controlling them in real-time could be implemented to enhance system performance over a wider range of conditions than originally designed for. In this paper, we propose an autonomous, reinforcement learning (RL) based, stormwater control system that aims to minimize flooding during storms. With this approach, an optimal control policy can be learned by letting an RL agent interact with the system in response to received reward signals. In comparison with a set of static control rules, RL shows superior performance on a wide range of artificial storm events. This demonstrates RL's ability to learn control actions based on observation and interaction, a key benefit for dynamic and ever-changing urban areas.
|
|
|
Christelle Pierkot, Sidonie Christophe, & Jean François Girres. (2019). Exploring multiplexing tools for co-visualization in crisis units. 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: Natural hazards can generate damages in large inhabited areas in a very short time period. Crisis managers must
plan interventions very quickly to facilitate the arrival of the first emergency. In a crisis unit, experts visualize
heterogeneous visual representations of spatio-temporal information, in order to facilitate decision-making,
based on various types of screens, i.e. laptops, tablets, or wall screens. Visualizing all this information at the
same time on the same interface would lead to cognitive overload. In this paper, we assume that it could be of
interest to provide innovative co-visualization models and tools, to bring hazard, geospatial and climate
information together, in a shared interface. We propose to explore spatial and temporal multiplexing tools within
a dedicated geovisualization environment, in order to help expert decision-making. The proposition is
implemented with the case study of a tsunami event in the Caribbean sea.
|
|
|
Christian Iasio, Ingrid Canovas, Elie Chevillot-Miot, & Tendry Randramialala. (2022). A New Approach to Structured Processing of Feedback for Discovering and Investigating Interconnections, Cascading Events and Disaster Chains. In Rob Grace, & Hossein Baharmand (Eds.), ISCRAM 2022 Conference Proceedings – 19th International Conference on Information Systems for Crisis Response and Management (pp. 285–298). Tarbes, France.
Abstract: Post-disaster information processing is relevant for the continuous improvement of operations and the reductionof risks. The current methodologies for post-disaster review suffer from several limitations, which reduce their use as a way of translating narrative in data for qualitative and quantitative analysis. Learning or effective knowledge sharing need a common formalism and method. Ontologies are the reference tool for structuring information in a “coded” data structure. Using the investigation of disaster management during the 2017 hurricane season in the French West Indies within the scope of the ANR “APRIL” project, this contribution introduces a methodology and a tool for providing a graphical representation of experiences for post-disaster review and lessons learning, based on a novel approach to case-based ontology development.
|
|
|
Christian Reuter, Gerhard Backfried, Marc-André Kaufhold, & Fabian Spahr. (2018). ISCRAM turns 15: A Trend Analysis of all ISCRAM-Papers 2004-2017. In Kees Boersma, & Brian Tomaszeski (Eds.), ISCRAM 2018 Conference Proceedings – 15th International Conference on Information Systems for Crisis Response and Management (pp. 445–458). Rochester, NY (USA): Rochester Institute of Technology.
Abstract: In 2004, Information Systems for Crisis Response and Management (ISCRAM) was a new area of research. Pioneering researchers from different continents and disciplines found fellowship at the first ISCRAM work-shop. Around the same time, the use of social media in crises was first recognized in academia. In 2018, the 15th ISCRAM conference will take place, which gives us the possibility to look back on what has already been achieved with regard to IT support in crises using social media. With this article, we examine trends and devel-opments with a specific focus on social media. We analyzed all papers published at previous ISCRAMs (n=1339). Our analysis shows that various platforms, the use of language and coverage of different types of disasters follow certain trends – most noticeably a dominance of Twitter, English and crises with large impacts such as hurricanes or earthquakes can be seen.
|
|
|
Christian Reuter, Marc-André Kaufhold, & René Steinfort. (2017). Rumors, Fake News and Social Bots in Conflicts and Emergencies: Towards a Model for Believability in 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. 583–591). Albi, France: Iscram.
Abstract: The use of social media is gaining more and more in importance in ordinary life, but also in conflicts and emer-gencies. The social big data, generated by users, is partially also used as a source for situation assessment, e.g. to receive pictures or to assess the general mood. However, the information's believability is hard to control and can deceive. Rumors, fake news and social bots are phenomenons that challenge the easy consumption of social media. To address this, our paper explores the believability of content in social media. Based on foundations of infor-mation quality we conducted a literature study to derive a three-level model for assessing believability. It summa-rizes existing assessment approaches, assessment criteria and related measures. On this basis, we describe several steps towards the development of an assessment approach that works across different types of social media.
|
|
|
Christian Siemen, Roberto dos Santos Rocha, Roelof P. van den Berg, Bernd Hellingrath, & João Porto de Albuquerque. (2017). Collaboration among Humanitarian Relief Organizations and Volunteer Technical Communities: Identifying Research Opportunities and Challenges through a Systematic Literature Review. 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. 1043–1054). Albi, France: Iscram.
Abstract: Collaboration is the foundation to strengthen disaster preparedness and for effective emergency response actions at all levels. Some studies have highlighted that remote volunteers, i.e., volunteers supported by Web 2.0 technologies, possess the potential to strengthen humanitarian relief organizations by offering information regarding disaster-affected people and infrastructure. Although studies have explored various aspects of this topic, none of those provided an overview of the state-of-the-art of researches on the collaboration among humanitarian organizations and communities of remote volunteers. With the aim of overcoming this gap, a systematic literature review was conducted on the existing research works. Therefore, the main contribution of this work lies in examining the state of research in this field and in identifying potential research gaps. The results show that most of the research works addresses the general domain of disaster management, whereas only few of them address the domain of humanitarian logistics.
|
|
|
Christina Tsouti, Eleni Ntzioni, Efstathia Tsarouchi, Dimitris Sakellariou, Marios Kotoulas, Christina Papadaskalopoulou, et al. (2022). Preparedness against Hazardous Events: A Novel Tool for Water Utilities. In Rob Grace, & Hossein Baharmand (Eds.), ISCRAM 2022 Conference Proceedings – 19th International Conference on Information Systems for Crisis Response and Management (pp. 185–199). Tarbes, France.
Abstract: Various terms and approaches currently exist on outlining the constituent components of crisis management as well as their interrelations, focusing mainly on the effective communication between the members of the crisis management unit. A gap emerges regarding a high-level and holistic approach on crisis management that will have the organization’s preparedness as its main pillar. In this work, crisis is organized into three macro-stages, i.e., pre-crisis, crisis, post-crisis. Preparedness is conceptualized as an overarching concept that frames an organization’s crisis management approach to reduce its vulnerability in a potential crisis. The study focuses on developing a high-level tool to enhance the preparedness of water utilities. The tool aims to serve as a holistic crisis management framework to support stakeholders in qualitatively assessing and improving their level of preparedness. The “Preparedness against hazardous events” tool was the result of this work, which was positively assessed through experts’ evaluation.
|
|
|
Christoph Amelunxen, & Janina Isabella Sander. (2019). Information collection using process visualisation in the risk management concept for emergency response. 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: Security-critical processes of emergency response are part of a complex system of people, organisation and
technology. They are often characterised by their own dynamics, interconnectedness and information deficits. In
addition, a wide variety of stakeholders, some from different organisations, work together, each specialising in a
specific area. In order to capture this (process-) knowledge in risk management, information from the experts is
necessary. However, experts are difficult to access, often separated locally, cost-intensive and usually have little
time (discussion-) capacity. A pictogram-based process visualisation was developed within the risk management
concept. The method could be validated within a European project in an expert workshop. This was done using
the example of a CBRN mass casualty incident. By using the methods presented, very good qualitative and
quantitative results can be achieved from the perspectives of various organisations and their experts. The limited
resource ?expert? is used optimally.
|
|
|
Christoph Lamers. (2022). Electronic Visualization for Situational Awareness in Control Rooms. In Rob Grace, & Hossein Baharmand (Eds.), ISCRAM 2022 Conference Proceedings – 19th International Conference on Information Systems for Crisis Response and Management (pp. 1008–1011). Tarbes, France.
Abstract: It is generally agreed in crisis management that a comprehensive visualization of the situation is crucial for an appropriate situational awareness of the staff personnel in control rooms. Therefore an expert group of fire officers in the German State North Rhine Westphalia developed a system for this purpose known as the “tactical wall”. The core of the system is a situation map of the relevant area with so-called tactical signs, i. e. defined graphic symbols for hazards, response units and tactical measures. Moreover, the assignment of response units to tactical sectors or staging areas as well as other relevant information such as the management organization is displayed at defined places within the wall. While the system was purely manual in its original version, a new digital version was recently developed. The user interfaces of this system are web-based and can by intuitively operated after a minor training effort.
|
|
|
Christopher W. Zobel, Milad Baghersad, & Yang Zhang. (2017). Calling 311: evaluating the performance of municipal services after disasters. 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. 164–172). Albi, France: Iscram.
Abstract: As part of a movement towards enabling smart cities, a growing number of urban areas in the USA, such as New York City, Boston, and Houston, have established 311 call centers to receive service requests from their citizens through a variety of platforms. In this paper, for the first time, we propose to leverage the large amount of data provided by these non-emergency service centers to help characterize their operational performance in the context of a natural disaster event. We subsequently develop a metric based on the number of open service requests, which can serve as the basis for comparing the relative performance of different departments across different disasters and in different geographic locations within a given urban area. We then test the applicability and usefulness of the approach using service request data collected from New York City's 311 service center.
|
|
|
Claire Laudy. (2017). Rumors detection on Social Media during Crisis Management. 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. 623–632). Albi, France: Iscram.
Abstract: Social Media monitoring has become a major issue in crisis and emergencies management. Indeed, social media may ease the sharing of information between citizens and Public Safety Organizations, but it also enables the rapid spreading of inaccurate information. As information is now provided and shared by anyone to anyone, information credibility is a major issue. We propose an approach to detect rumor in social media. This paper describes our work on semantic graph based information fusion, enhanced with uncertainty management capabilities. The uncertainty management capability enables managing the dierent level of credibility of actors of an emergency (dierent PSO oÿcers and citizens). Functions for information synthesis, conflicting information detection and information evaluation were developed and test during experimentation campaigns. The synthesis and conflicting information detection functionalities are very welcome by end-users. However, the uncertainty management is a combinatorial approach which remains a limitation for use with large amount of information.
|
|
|
Clara Grimes, Mihoko Sakurai, Vasileios Latinos, & Tim A Majchrzak. (2017). Co-creating Communication Approaches for Resilient Cities in Europe: the Case of the EU Project Smart Mature Resilience. 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. 353–362). Albi, France: Iscram.
Abstract: Cities face a wide range of risks. Potential threats range from natural disasters and the (relatively slow) environmental change, to man-made issues like extremism. To overcome such threats, cities ought to be resilient, capable of resisting problems, of adapting to new situations, and overcoming crises. Effective communication is particularly crucial for a resilient city. Rather than trusting that relevant stakeholders, municipal staff and citizens will intuitively communicate in the ideal way, cities should see communication as a strategic aspect of their resilience development. Thus, how resilient cities communicate should be strategically managed. In this paper, we present immediate results from an ongoing European project called Smart Mature Resilience. In this project, we work with seven cities towards the ultimate goal of developing a Resilience Management Guideline for all European cities. Moreover, we intend to set up a resilience backbone in Europe, which will be driven by effective communication between cities.
|
|
|
Clara Le Duff, Jean-Philippe Gitto, Julien Jeany, Raphaël Falco, Matthieu Lauras, & Frederick Benaben. (2022). A Physics-based Approach to Evaluate Crisis Impacts on Project Management. In Rob Grace, & Hossein Baharmand (Eds.), ISCRAM 2022 Conference Proceedings – 19th International Conference on Information Systems for Crisis Response and Management (pp. 134–143). Tarbes, France.
Abstract: Project management has become a standard in business. Unfortunately, the projects as well as companies are increasingly subject to major disruptions. In this context, it is of prime importance to have the ability to manage the risks inherent to these projects to best achieve their objectives. The existing approaches of crisis management in the literature no longer seem to be adapted to this new normality. The future of research lies in a more systematic crisis assessment and a better conceptualization of the uncertainty associated with risks. It is necessary to rely on the collection of heterogeneous data in order to maximize the understanding of the project environment and to find a way that best describes and visualizes the influence of crises on the project management processes. This article uses the POD approach and applies it in the context of project management to address these issues.
|
|
|
Claudio Arbib, Davide Arcelli, Julie Dugdale, Mahyar Tourchi Moghaddam, & Henry Muccini. (2019). Real-time Emergency Response through Performant IoT Architectures. 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 describes the design of an Internet of Things (IoT) system for building evacuation. There are two main
design decisions for such systems: i) specifying the platform on which the IoT intelligent components should be
located; and ii) establishing the level of collaboration among the components. For safety-critical systems, such as
evacuation, real-time performance and evacuation time are critical. The approach aims to minimize computational
and evacuation delays and uses Queuing Network (QN) models. The approach was tested, by computer simulation,
on a real exhibition venue in Alan Turing Building, Italy, that has 34 sets of IoT sensors and actuators. Experiments
were performed that tested the effect of segmenting the physical space into different sized virtual cubes. Experiments
were also conducted concerning the distribution of the software architecture. The results show that using centralized
architectural pattern with a segmentation of the space into large cubes is the only feasible solution.
|
|
|
Claudio Paliotta, Klaus Ening, & Sigurd Mørkved Albrektsen. (2021). Micro indoor-drones (MINs) for localization of first responders. 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. 881–889). Blacksburg, VA (USA): Virginia Tech.
Abstract: In this paper, we describe our approach to the localization in GNSS-denied and risky unknown environments offirst responders (FRs). The INGENIOUS project is an EU funded project which is developing a new integratedtoolkit to support the operations of FRs. The micro indoor-drones (MINs) developed within the INGENIOUSproject represent a component of the toolkit which will support the localization of FRs in search-and-rescue (SAR)operations. In this paper, the concept behind the MINs and the current achievements are illustrated.
|
|
|
Cody Buntain, Richard Mccreadie, & Ian Soboroff. (2022). Incident Streams 2021 Off the Deep End: Deeper Annotations and Evaluations in Twitter. In Rob Grace, & Hossein Baharmand (Eds.), ISCRAM 2022 Conference Proceedings – 19th International Conference on Information Systems for Crisis Response and Management (pp. 584–604). Tarbes, France.
Abstract: This paper summarizes the final year of the four-year Text REtrieval Conference Incident Streams track (TREC-IS), which has produced a large dataset comprising 136,263 annotated tweets, spanning 98 crisis events. Goals of this final year were twofold: 1) to add new categories for assessing messages, with a focus on characterizing the audience, author, and images associated with these messages, and 2) to enlarge the TREC-IS dataset with new events, with an emphasis of deeper pools for sampling. Beyond these two goals, TREC-IS has nearly doubled the number of annotated messages per event for the 26 crises introduced in 2021 and has released a new parallel dataset of 312,546 images associated with crisis content – with 7,297 tweets having annotations about their embedded images. Our analyses of this new crisis data yields new insights about the context of a tweet; e.g., messages intended for a local audience and those that contain images of weather forecasts and infographics have higher than average assessments of priority but are relatively rare. Tweets containing images, however, have higher perceived priorities than tweets without images. Moving to deeper pools, while tending to lower classification performance, also does not generally impact performance rankings or alter distributions of information-types. We end this paper with a discussion of these datasets, analyses, their implications, and how they contribute both new data and insights to the broader crisis informatics community.
|
|
|
Cody Buntain, Richard Mccreadie, & Ian Soboroff. (2021). Incident Streams 2020: TRECIS in the Time of COVID-19. 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. 621–639). Blacksburg, VA (USA): Virginia Tech.
Abstract: Between 2018 and 2019, the Incident Streams track (TREC-IS) has developed standard approaches for classifying the types and criticality of information shared in online social spaces during crises, but the introduction of SARS-CoV-2 has shifted the landscape of online crises substantially. While prior editions of TREC-IS have lacked data on large-scale public-health emergencies as these events are exceedingly rare, COVID-19 has introduced an over-abundance of potential data, and significant open questions remain about how existing approaches to crisis informatics and datasets built on other emergencies adapt to this new context. This paper describes how the 2020 edition of TREC-IS has addressed these dual issues by introducing a new COVID-19-specific task for evaluating generalization of existing COVID-19 annotation and system performance to this new context, applied to 11 regions across the globe. TREC-IS has also continued expanding its set of target crises, adding 29 new events and expanding the collection of event types to include explosions, fires, and general storms, making for a total of 9 event types in addition to the new COVID-19 events. Across these events, TREC-IS has made available 478,110 COVID-related messages and 282,444 crisis-related messages for participant systems to analyze, of which 14,835 COVID-related and 19,784 crisis-related messages have been manually annotated. Analyses of these new datasets and participant systems demonstrate first that both the distributions of information type and priority of information vary between general crises and COVID-19-related discussion. Secondly, despite these differences, results suggest leveraging general crisis data in the COVID-19 context improves performance over baselines. Using these results, we provide guidance on which information types appear most consistent between general crises and COVID-19.
|
|
|
Congcong Wang, Paul Nulty, & David Lillis. (2021). Crisis Domain Adaptation Using Sequence-to-Sequence Transformers. 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. 655–666). Blacksburg, VA (USA): Virginia Tech.
Abstract: User-generated content (UGC) on social media can act as a key source of information for emergency responders incrisis situations. However, due to the volume concerned, computational techniques are needed to effectively filter and prioritise this content as it arises during emerging events. In the literature, these techniques are trained using annotated content from previous crises. In this paper, we investigate how this prior knowledge can be best leveraged for new crises by examining the extent to which crisis events of a similar type are more suitable for adaptation tonew events (cross-domain adaptation). Given the recent successes of transformers in various language processing tasks, we propose CAST: an approach for Crisis domain Adaptation leveraging Sequence-to-sequence Transformers. We evaluate CAST using two major crisis-related message classification datasets. Our experiments show that ourCAST-based best run without using any target data achieves the state of the art performance in both in-domain and cross-domain contexts. Moreover, CAST is particularly effective in one-to-one cross-domain adaptation when trained with a larger language model. In many-to-one adaptation where multiple crises are jointly used as the source domain, CAST further improves its performance. In addition, we find that more similar events are more likely to bring better adaptation performance whereas fine-tuning using dissimilar events does not help for adaptation. To aid reproducibility, we open source our code to the community.
|
|
|
Congcong Wang, Paul Nulty, & David Lillis. (2021). Transformer-based Multi-task Learning for Disaster Tweet Categorisation. 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. 705–718). Blacksburg, VA (USA): Virginia Tech.
Abstract: Social media has enabled people to circulate information in a timely fashion, thus motivating people to post messages seeking help during crisis situations. These messages can contribute to the situational awareness of emergency responders, who have a need for them to be categorised according to information types (i.e. the type of aid services the messages are requesting). We introduce a transformer-based multi-task learning (MTL) technique for classifying information types and estimating the priority of these messages. We evaluate the effectiveness of our approach with a variety of metrics by submitting runs to the TREC Incident Streams (IS) track: a research initiative specifically designed for disaster tweet classification and prioritisation. The results demonstrate that our approach achieves competitive performance in most metrics as compared to other participating runs. Subsequently, we find that an ensemble approach combining disparate transformer encoders within our approach helps to improve the overall effectiveness to a significant extent, achieving state-of-the-art performance in almost every metric. We make the code publicly available so that our work can be reproduced and used as a baseline for the community for future work in this domain.
|
|
|
Cornelius Dold, Christopher Munschauer, & Ompe Aimé Mudimu. (2020). Real-Life Exercises as a Tool in Security Research and Civil Protection – Options for Data Collections. 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. 244–250). Blacksburg, VA (USA): Virginia Tech.
Abstract: A real-life exercise is a scientific method used by the TH Köln to generate data sets of new technologies and operational concepts derived from research projects. The Institute of Rescue Engineering and Civil Protection (German acronym: IRG) uses a real-time locating system (RTLS), video surveillance, observers and a mass casualty incident benchmark to generate motion profiles, information flows and information on the quality of care. In this practitioner paper these different methods will be discussed and the combination of different data is described. Furthermore, an outlook is given on the extent to which the method will be improved and expand-ed in the future. Concluding it can be said that the combination of all collected data is essential for the evalua-tion of a real-life exercise in security research or civil protection.
|
|