Abstract: In this paper, we develop a model with neural networks to localize events using microblogging data. Localization is the task of finding the location of an event and can be done by discovering event signatures in microblogging data. We use the deep learning methodology of Bi-directional Long Short-Term Memory (Bi-LSTM) to learn event signatures. We propose a methodology for labeling the Twitter date for use in Bi-LSTM However, there might not be enough data available to train the Bi-LSTM and learn the event signatures. Hence, the data is augmented using generative adversarial networks (GAN). Finally, we combine event signatures at different temporal and spatial granularity to improve the accuracy of event localization. We use microblogging data collected from Twitter to evaluate our model and compare it with other baseline methods.

Abstract: Twitter is widely being used by researchers to understand human behavior, e.g. how people behave when an event occurs and how it changes their microblogging pattern. The changing microblogging behavior can have an important application in the form of detecting events. However, the Twitter data that is available has limitations in it has incomplete and noisy information and has irregular samples. In this paper we create a model, calledTwitter Behavior Agent-Based Model (TBAM)to simulate Twitter pattern and behavior using Agent-Based Modeling(ABM). The generated data can be used in place or to complement the real-world data and improve the accuracy of event detection. We confirm the validity of our model by comparing it with real data collected from Twitter

Abstract: In this paper we consider how emergency response organizations utilize available social media technologies to communicate with the public in emergencies and to potentially collect valuable information using the public as sources of information on the ground. We discuss the use of public social media tools from the emergency management professionals. viewpoint with a particular focus on the use of Twitter. Little research has investigated Twitter usage in crisis situations from an organizational perspective. This paper contributes to our understanding of organizational innovation, risk communication, and technology adoption by emergency management. An in-depth case study of Public Information Officers of the Los Angeles Fire Department highlights the importance of the information evangelist within emergency management organizations and details the challenges those organizations face with an engagement with social media and Twitter. This article provides insights into practices and challenges of new media implementation for crisis and risk management organizations.

Abstract: Social media are increasingly becoming a source for event-based early warning systems in the sense that they can help to detect natural disasters and support crisis management during or after disasters. In this work-in-progress paper we study the problems of analyzing multilingual twitter feeds for emergency events. The present work focuses on English as “lingua franca” and on under-resourced Mediterranean languages in endangered zones, particularly Turkey, Greece, and Romania Generally, as local civil protection authorities and the population are likely to respond in their native language. We investigated ten earthquake events and defined four language-specific classifiers that can be used to detect earthquakes by filtering out irrelevant messages that do not relate to the event. The final goal is to extend this work to more Mediterranean languages and to classify and extract relevant information from tweets, translating the main keywords into English. Preliminary results indicate that such a filter has the potential to confirm forecast parameters of tsunami affecting coastal areas where no tide gauges exist and could be integrated into seismographic sensor networks. © 2012 ISCRAM.

Abstract: This research presents a centralized boundary object website and mobile app focused on allowing participants to participate in developing an early warning system through space weather and the beauty of the aurora borealis. Because of the beauty and majesty of auroral activity, people will seek information about when and where these unpredictable events occur. This activity, commonly referred to as nowcasting, can be combined with scientific data collected from observatories and satellites and serve as an early warning system with potentially far greater accuracy and timeliness than the current state of the art. We believe that long-term engagement with a citizen science tool will help bridge the many social worlds surrounding the aurora borealis and lead to the development of an early warning system that may correlate the visibility of the northern lights to violent space weather. We hope this will lead to other real time crowdsourced early warning systems in the future.

Abstract: To this date, research on crisis informatics has focused on the detection of trust in Twitter data through the use of message structure, sentiment, propagation and author. Little research has examined the effects of perceived emotion of these messages in the crisis response domain. Toward detecting useful messages in case of crisis, we examine perceived emotions of these messages and how the different emotions affect the perceived usefulness and trustworthiness. Our analysis is carried out on two datasets gathered from Twitter concerning hurricane Sandy in 2012 and the Boston Bombing 2013. The results indicate that there is a significant difference in the perceived emotions that contribute towards the perceived trustworthiness and usefulness. This could have impacts on how messages from social media data are analyzed for use in crisis response.

Abstract: Social Media and micro-blogging is being used during crisis events to provide live up-to-date information as events evolve (before, during and after). Messages are posted by citizens or public officials. To understand the effectiveness of these messages, we examined the content of geo-located Twitter messages (“tweets”) sent during the Moore, Oklahoma tornado of May 20th, 2013 (+/-1day) to explore the spatial and temporal relationships of real-time reactions of the general public. We found a clear transition of topics during each stage of the tornado event. Twitter was useful for posting and retrieving updates, reconstructing the sequence of events as well as capturing people's reactions leading up to, during and after the tornado. A long-term goal for the research reported here is to provide insights to forecasters and emergency response personnel concerning the impact of warnings and other advisory messages.

Abstract: Microblogging services like Twitter afford opportunities for real time determination of situation awareness during crises as people report, via their statuses, information about events on the ground. An important component of the information included in a tweet are mentions of place names that may be sites of damage, injuries, or relief efforts. Methods for extracting these place names and determining the actual location being referenced are an essential part of the suite of tools required for automated extraction of situation awareness from tweets. Extracting and disambiguating place name mentions from text have been areas of extensive research. Twitter, however, presents challenges given the 140 character restriction on status and the informal, abbreviated language that are a norm in this communication channel. Named entity recognizers, which are dependent on labeled training data, may not be useful in this medium for extracting location mentions because the typical training domains for these taggers are absent the noise found in Twitter statuses. Additionally, the contextual information that is necessary for disambiguating place names is not always present. In this paper, we demonstrate a new technique, RapidGeo, for extracting and disambiguating place names from a location specific Twitter feed using an unsupervised technique for tagging location mentions and relying on the known geographic context of the feed for disambiguation. Our location tagging technique performs much better than an off-the-shelf named entity recognizer and we achieve reasonable precision in disambiguating extracted place names. We argue that such fast, high precision, unsupervised approaches are needed when important, actionable information is required from noisy data sources such as Twitter. © 2012 ISCRAM.

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: 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.

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: In this paper we present findings from interviews conducted with representatives from large international disaster response organizations concerning their use of social media data in crisis response. We present findings in which the barriers to use by responding organizations have gone beyond simple discussions of trustworthiness to that of more operational issues rather than mere data quality. We argue that the landscape of the use of microblogged data in crisis response is varied, with pockets of use and acceptance among organizations. We found that microblogged data is useful to responders in situations where information is limited, such as at the beginning of an emergency response effort, and when the risks of ignoring an accurate response outweigh the risks of acting on an incorrect one. In some situations, such as search and rescue operations, microblogged data may never meet the standards of quality required. In others, such as resource and supply management, microblogging data could be useful as long as it is appropriately verified and classified.