Twitter* and Online Fraud Prevention

*At the time of this project Twitter was not known as X. We have thus used the original name in the text below.

Research summary

In recent years, social media has been increasingly used for public engagement and to aid in police investigations. Twitter alone has over 368 million monthly active users worldwide with over 500 million daily posts. Most people use the network to get the latest news, entertainment or to express their opinions. Given the popularity of social media platforms and the general public’s engagement with them, it is not surprising that public bodies and government agencies across the world commonly employ them to communicate with the public through verified user accounts. 

However, to the best of our knowledge, research concerning the use of social media by crime reduction stakeholders and their interactions with users is limited. In terms of crime prevention, the majority of research which uses social media data does so to predict crime trends using machine learning, mostly by examining the content or sentiment of tweets over time and correlating this with time-series data concerned with urban crime and other factors such as unemployment or weather conditions (Aghababaei and Makrehchi, 2018; Chen et. Al. 2015; Gerber, 2014). In terms of online crime, previous research has generally focused on the detection of social media bots (software that acts autonomously) and fake user profiles (Albahar, 2019; Gurajala et al. 2015), or other crimes prominent on social media, such as hate crime (Müller and Schwarz, 2020; Alnazzawi, 2022) and cyber bulling (Al-Garadi et al, 2016). 

While the above work is valuable, it does not focus on: 1) how crime reduction stakeholders use social media for the purposes of reducing crime such as online fraud: 2) how social media users interact with the accounts of stakeholders; and, 3) the potential utility of the content of any interactions that take place (e.g. do responses provide useful intelligence that the police might exploit). Perhaps not surprisingly then, previous research has explicitly called for an evidence base of how the policing community uses social media and “what works” to inform the public (Nikolovska, Johnson and Ekbolm, 2020; Fernandez, Dickinson and Alani, 2017).  

This project seeks to address these research gaps through: 

1. A thematic analysis of the content of tweets posted by crime reduction stakeholders (e.g do they post safety advice, reporting advice, information on new forms of fraud etc).
2. An analysis of user interaction with such content (e.g. likes, retweets, shares) to assess ‘what works’ in terms of increasing the coverage and impact of prevention campaigns and online community outreach for online fraud. 
3. A thematic analysis of the content of tweets posted by users in response to stakeholder tweets (e.g. is it general service feedback, reporting issues, requests for advice on how to reduce risk, descriptions of victimisation experiences or the modus operandi of emerging forms of online fraud) to assess the utility of such information for stakeholders and the reduction of online fraud.
4. Understanding if crime reduction stakeholders have a strategy for the use of social media, how they use it, and what the intended outcomes of posts or campaigns of them are.

The findings of this project would be used to develop guidance on how the police and other crime reduction stakeholders might better use Twitter (and potentially other social media networks) to prevent online fraud, and (where this fails) to improve the reporting of it.