The challenges and concerns of using big data to understand cybercrime

Introduction

The proliferation of computers, mobile devices, and the internet have revolutionised all aspects of modem life. Traditional boundaries of space and time have been rendered meaningless, as individuals can interact in real time through text or instant messaging, as well as asynchronously via email, forums, and other forms of computer-mediated communications (Holt & Bossier, 2016; Newman & Clarke, 2003; Wall, 2001). Individuals are also no longer limited to either written or voice messaging, as decentralised social networking sites like Facebook, Instagram, and Snapchat enable users to share content using any combination of text, images, and video (Holt & Bossier, 2016). The internet has also changed the landscape of financial transactions and commerce, making it possible for people to buy goods and services from vendors around the world. Retailers maintain customers’ financial data in large databases, enabling payments to be made instantly via various currencies (Holt & Bossier, 2016).

Given the relative ease with which individuals can access the internet across various devices, most nations are now dependent on technology for critical services (Andress & Winterfeld, 2013; Holt & Bossier, 2016; Newman & Clarke, 2003). These benefits have also created unparalleled opportunities for individuals to misuse online resources to engage in a variety of deviant and criminal acts (Holt & Bossier, 2016). Anew lexicon has developed to define the scope of these offenses, particularly the phrase cybercrime, recognising the use of the internet and/ or internet-enabled devices to perform acts of crime and deviance (Fumell, 2002; Wall, 2001). Traditional crimes that occur in real world environments have been enhanced by technology, including prostitution (Cunningham & Kendall, 2010; Holt & Blevins, 2007), fraud (Cross, 2015; Wall, 2004), stalking and harassment (Bocij, 2004; Choi & Lee, 2017; Choi, Lee, & Lee, 2017), terrorism (Britz, 2010; Weimann, 2005), piracy (Higgins & Marcum, 2011), and child sexual offenses (Jenkins, 2001; Krone, 2004). New forms of crime have also developed as a function of technology, including unauthorised access of computers and the dissemination of malicious software (Bossier & Holt, 2009; Jordan & Taylor, 1998).

The rise of cybercrime presents a unique challenge for academic researchers and policy-makers, as the number of calls for service for these offenses, the composition of offender populations, and victim characteristics are not well explicated in official data sources such as the Uniform Crime Report in the US, or the British Crime Survey in England and Wales (Brenner, 2009; Wall, 2007). Additionally, large-scale, nationally representative self-report studies of deviance and crime do not include measures for cybercriminality, making it difficult to understand the predictors and pathways to offending (Holt & Bossier, 2016).

Some researchers have begun to use alternative data sources as a means to understand cybercrime, specifically text and visual information posted by individuals in clear-text through various online data sources, including websites, social media feeds, and computer-mediated communications platforms (see Holt, 2017; Silverman, 2013 for reviews). Offenders discuss then motivations, methods, and target preferences on forums. Facebook, Twitter, and other sites due to the degree of anonymity afforded by the internet. In addition, some platforms exist solely to facilitate overt criminality, such as illicit markets operating to sell personal information (e.g., Franklin. Paxson, Penig, & Savage, 2007; Holt & Lampke, 2010; Hutchings & Holt, 2015), hacking tools (Dupont, Côté, Boutin, & Fernandez, 2017; Holt, 2013b), drags (e.g., Barratt, 2012; Decary-Hetu & Quessy-Dore, 2017), and sexual services (Horswill & Weitzer, 2018; Kosloski, Bontrager-Ryon, & Roe-Sepowitz, 2017). These sources provide deep insights at the individual and aggregate level to understand both offender perspectives and potential targets for victimisation.

To date, most researchers have used online data collection methods to examine cybercrimes using both qualitative and quantitative models (see Holt & Bossier, 2016 for review). While cross-sectional and longitudinal studies are useful for examining relationships between predetermined variables, these analyses are incapable of interpreting the associations found within individualised transactions established in cyberspace (Song, Song, An, Hayman, & Woo, 2014). The quantity of data available through online sources, including years of text and images, could enable big data mining and analysis techniques to assess the practices of online offenders and then potential victims (Chen, Mao, & Liu, 2014; Clark, 2017; Song, Song. & Lee, 2018; Williams, Вшпар, & Sloan, 2017). While big data has the ability to illustrate various trends that can be later examined to support and/or challenge wider assumptions, they do not make any inferences themselves (Smith, Moses, & Chan, 2017; Song et al., 2018). The benefits of analysing big data rests in its ability to consolidate a larger volume of data and confinn the thoughts of a host of participants, lending itself to more accurate predictions (Song et al., 2018; Williams et al., 2017). Additionally, though big data research is commonplace among certain disciplines (e.g., musing, engineering, business, and medicine), it has not been used with much consistency in criminal justice (see Smith et al., 2017; Williams et al., 2017).

Despite the various challenges and limitations of big data, it is a burgeoning phenomenon that demands more attention and consideration within inquiries that investigate cybercrime and cybersecurity. This chapter will examine the challenges and concerns of using big data within cybercrime research, beginning with an introduction of big data and its defining characteristics. Then, we consider why

Using big data to understand cybercrime 87 and how big data could be applied uniquely to crime, as well as perspectives on how big data could prevent the emergence of modem crime problems. The chapter will conclude with a discussion on the risks and implications posed by using big data to examine cybercrime offending and victimisation.