Introduction: dealing with information in crisis management
Today, societies face many potential threats that can turn into crisis situations. Crises (emergencies) upset society, and put its critical infrastructures under stress (Quarantelli 1998; Comfort et al. 2010).1 Once a crisis occurs organizations, both public and private, are supposed to âfightâ the crisis and form coalitions with other agencies and local communities. Since crises are often characterized by multiple causes, ambiguity of effects and various means of resolution, as well as by a shared belief that decisions must be made swiftly (Pearson and Clair 1998; Van der Vegt et al. 2015), information management is a vital component of preparedness, response and relief. An adequate and effective information management that supports crisis organizations requires processes to collect, analyze and share information about the crisis situation, and about the coordination between the responding organizations. When a crisis occurs, information managers start to collect and produce standard information products to support the coordination of the response operation (Comfort et al. 2004; Oh et al. 2013).
In addition to the data collected, shared, analyzed and used by official organizations, administrations and mainstream media, citizens inform themselves and others about crisis situations through social media platforms, generating bottom-up information networks (Palen 2008; Hughes and Palen 2009; Yates and Paquette 2011). All these actions contribute to the âexplosionâ in the amount of data and information at times of disasters, which is a challenge for responding organizations to deal with. For example, because crisis information may become outdated soon as crisis conditions change, crisis response needs the management of information flows and networks to build an effective crisis response organization (Pan et al. 2012). Crisis responders then rely on traditional information systems such as enterprise resource systems, but since digital data are practically ubiquitous, the emerging information networks form potentially useful additional sources for the organization of the crisis response. Together, they create a crisis information ecology of dynamic information streams (Turoff et al. 2004; Van de Walle et al. 2009). Information ecology traditionally refers to the total information environment of organizations (Davenport and Prusak 1997) â to understand the characteristics of this ecology is of crucial importance to grasp how people really use information, how they search for it, modify it, share it, or even ignore it. Crisis information management implies that data can be translated into âactionableâ information to increase the quality of the crisis response (Boersma et al. 2012; Wolbers and Boersma 2013). In a crisis situation the information ecology leads to a crisis information paradox: on the one hand the (governmental) responding organizations and administrations want to stay in control by harvesting and integrating the various and heterogeneous data sources in their information management systems, on the other hand the complex nature of the information ecology make an authoritarian response structure virtually impossible.
With the increased availability of data for effective crisis response, new challenges are added to the burden of crisis management. There are serious concerns related to the (lack of) information standards and accountability mechanisms (Turoff 2002), information overload (Hiltz and Plotnick 2013), the lack of interoperability between the information and communication technologies used by the first responders and the communication sources used by citizens (Truptil et al. 2008), and underdeveloped (big) data analytical skills by the users of crisis information. At the same time, crises, disasters and social disruptions are seen as opportunity windows to create legitimacy to collect and analyze citizensâ data on a large scale (Fonio et al. 2007). In other words, the use of crisis information systems, i.e., networks of hardware and software, to create, collect, filter, process and distribute data is not neutral, but related to the way crisis information management is organized and legitimized.
The big data debate in crisis and disaster management
Increasingly, crisis information management includes the processing and use of big data by (governmental) responding organizations in order to try to control the crisis situation. Big data refers to a quantitative increase of the size of the datasets that can be used for analytical purposes by a wide range of actors, including academics, marketers, governmental bodies, educational institutions and â in the context of this book â crisis managers (boyd and Crawford 2012; Shelton et al. 2014). One of the most widely accepted ways to describe big data is the âthree Vsâ (volume, variety and velocity) of information (McAfee et al. 2012). âVolumeâ refers to the generation and collection of data, and implies that the data volume becomes increasingly larger. âVelocityâ addresses the timeliness of the data, and the speed of data collection, analysis and use to maximize its utility; finally, âVarietyâ indicates the various types of data, including semi-structured, unstructured, validated and unvalidated, raw and analyzed data and its technical sources, such as audio, video, webpage and text (Mayer-Schönberger and Cukier 2013; Chen et al. 2014). Potentially the use of big data will change the way responding organizations make sense of the crisis situations, respond to it and make decisions concerning the crisis response.
For example, a serious challenge at times of crisis is to create a âcommon operational pictureâ of the situation and of the actions and interactions of others involved in the crisis management (Wolbers and Boersma 2013). Crisis managers can use big data analytics to create improved operational pictures (Wukich 2015). Another example is the use of social media data by crisis management organizations as part of early warning systems (Culotta 2010), and for crowd control and monitoring (Trottier and Schneider 2012; Boersma 2013; Procter et al. 2013). There is growing evidence that social media data can contribute to a better understanding of the situation and eventually to a more adequate and robust crisis management (Yin et al. 2012; Cassa et al. 2013). The use of social media data in crisis management, its intended and unintended consequences, is a central issue in the first part of this book (Chapters 2, 3 and 4). Because of the promising character of social media data governmental administrations, private organizations and non-governmental organizations invest a lot in crisis management information systems that can harvest valuable data from social media sources. For example, Twitcident is a tool used by professionals in emergency control rooms to follow what (relevant) data citizens post on Twitter for the purpose of maintaining security in urban environments (Boersma et al. 2016).
The use of big data for any purpose should not be taken for granted as it requires adequate data and information management (Pries and Dunnigan 2015). Databases can indeed generate patterns that have predictive power for the crisis operations but not necessarily and automatically explanatory power (Andrejevic 2014). It is the extraction of structured data from unstructured inputs that is the most challenging and the biggest gap in the understanding of those who want to use big data in the context of crisis response (Castillo 2016). The availability of big crisis data does not always entail, let alone guarantee, effective crisis management.
However, Floridi (2012) argues that becoming data-richer by the day cannot be perceived as a fundamental problem per se. Big data undoubtedly represents an opportunity in disaster management, especially since âdigital humanitariansâ appeared on the scene. From the 2010 Haiti earthquake onward, disaster response has been redefined by new players, namely digital volunteers who have supported search and rescue efforts through, for instance, the generation of maps or the interpretation of large amounts of data (Mulder et al. 2016). Digital humanitarians â as they are labeled â form a âcrowdâ that provides various services, such as building situational awareness from social media or generating maps, while using information and communication technology (Link et al. 2014). Digital humanitarians have played a vital role in verifying the accuracy of information shared in social media during crises and, in some cases, they have actively shaped disaster response in the aftermath of a major event by helping first respondersâ organizations (Burns 2014).
The rise of big crisis data has been explored in the context of humanitarian response, in particular during, or in the aftermath of, a natural disaster (Meier 2015; Castillo 2016). Increasingly, a sheer amount of data is generated through social media during crises: when a major disaster strikes, a âdigital nervous systemâ (Meier 2015: 27) reacts through various synapses encapsulated in various forms of communication, from tweets to pictures posted on social media. While, in this specific context, the expression âbig crisis dataâ does not have a negative connotation but instead refers to data generated by affected communities and used for the purpose of helping them, it is worth noting that a disaster can turn into a âbig data crisisâ if first response organizations do not have the capacity to deal with potential valuable information shared in social media. As emphasized by the International Federation of Red Cross et al. in 2005 âpeople need information as much as water, food medicine or shelter. Information can save lives, livelihood and resources. Information bestows power.â Therefore, in current practices of disaster management, it is essential to ensure a proper use of social media during crisis to respond to the information needs of the communities affected by disasters.
It means that the use of big data at times of crisis (and the outcome of the digital humanitariansâ actions for that matter) is not without problems. Like any hype in information and communication technology it asks for a critical analysis: it can trigger processes of change, but also easily can become an empty promise (Meijer et al. 2009). A real epistemological problem with big data, according to Floridi, is detecting small and meaningful patterns. This is of particular relevance in the field of crisis management and raises questions that seem to remain unsolved, such as to what extent real-time big crisis data can enhance disaster response instead of turning into a big data crisis due to the challenges of working with new data sources. Hence, the debate on the use of big data is concerned with methods used to make sense of data (namely, detecting meaningful small patterns) and decisions made upon the interpretation of patterns. Big crisis data is subject to interpretation and bias like any other data sources (boyd and Crawford 2012). In addition, humanitarianism has been critiqued as a social relation that often privileges people from the global North: data and technologies often reify social and power relations, worldviews and epistemologies (Elwood and Leszczynski 2013; Burns 2015).
In sum: big crisis data should not be considered as a magic bullet which can save lives just because they are available.
Surveillance crisis management: the intended and unintended consequences of big data in use
Whereas in the creation of common operational pictures the use of crisis data from social media and other data sources is promising but problematic in itself for various practical and more fundamental reasons (because of the reasons addressed above), in this edited volume we are in particularly interested in the surveillance aspect of crisis management. We believe the surveillance debate is significant for the crisis and disaster studies. The surveillance âlensâ is a powerful âempirical windowâ through which we witness how people and their data doubles (i.e., the online identities or classifications that represent the individual to which they are attached; see Lyon 2007) are being monitored and controlled (Jenness et al. 2007) at times of disasters â and as a consequence of disaster relief.
In disaster response, surveillance practices are used for different purposes and in different phases. Currently the big data debate in disaster management cannot be disentangled from the role of digital humanitarians who seem to have made good use of surveillance practices (e.g., data mining) on the internet. These practices resonate with the concept of âlateral surveillanceâ as defined by Andrejevic (2002): the use of surveillance tools by individuals rather than by institutions to keep track of each other for several purposes. One could argue that digital humanitarians practice lateral surveillance for humanitarian and crisis management purposes. For instance, surveillance has taken the form of automatic classification of tweets or mapping geo-tagged information. These practices have been explored through lenses which are different from the dystopian views sometimes embedded into surveillance studies. At the same time, data collection, especially of people affected by disasters through different means, is also considered as a routine practice in order to assist individuals and communities. The dimension of control, however, is often overlooked in the literature of crisis management due to the positive connotation of control for assessing needs, helping people, and counting human and economic losses.
The surveillance lens helps us to understand how crisis management has become an integral part of what has been called the âsurveillance societyâ (Gandy 1989; Wood et al. 2006; Ball et al. 2012). Surveillance refers to the rational modernistic thinking: âany collection and processing of personal data, whether identifiable or not, for the purposes of influencing or managing those whose data have been garneredâ (Lyon 2001). Surveillance is a consequence of processes of modernity (Giddens 1985) and has become an inherent part of our network societies (Castells 2001). Although the state and state agencies have been playing a major role in surveillance societies (Haggerty and Samatas 2010; Wagenaar and Boersma 2008; Webster et al. 2012; Boersma et al. 2014), surveillance is about much more than state control. Haggerty and Samatas define surveillance as an activity that involves âassorted forms of monitoring, typically for the ultimate purpose of intervening in the worldâ (2010, p. 2). The use of computerized systems enables electronic forms of surveillance, not just because electronic databases made it easy to store huge amounts of personal data, but because it has changed surveillance practices.
The speed of data flows has increased, databases became decentralized and easily accessible, and individuals more easily traced. The internet has enabled a global networked form of surveillance (Fuchs et al. 2011). It has led to datafication as a new paradigm in science and society (Van...
