DAPM: Detecting and Preventing Mass-Marketing Fraud (MMF)

欺诈可以广义地定义为用来获得相对于另一个人或组织的不诚实的优势(通常是财务优势)的诡计。大规模营销欺诈(MMF)是一种利用大众传播技术(如电子邮件、即时消息、群发邮件、社交网站、电话营销)骗取他人钱财的欺诈行为。这是一种针对大多数国家受害者的骗局。尼日利亚的419骗局就是一个例子。在这个骗局中，诈骗者要求预付费用，并承诺受害者将以微不足道的努力收回大笔资金。然而，并不是所有的大众营销诈骗都以赚取大笔金钱的承诺来欺骗受害者。例如，在爱情骗局中，罪犯假装与受害者建立了浪漫的关系，后来要求金钱帮助他们，特别是在危机中。在这起慈善骗局中，受害者相信他们是在把钱捐给真正的慈善机构。拟议中的项目将开发新的技术来检测和预防MMF。通过其多学科方法和密切关注于与一系列项目合作伙伴共同设计解决方案，并在现场MMF检测环境中对其进行野外测试，该项目不仅将导致对MMF解剖的新的科学理解，而且还将产生可构成打击此类欺诈的实际干预措施基础的工具和技术。与学术之外的合作伙伴合作对这个项目的成功至关重要。多年来，各种类型的组织一直在努力检测和预防MMF(通常是在竖井中)-一些方法似乎有些有效。尽管如此，受害者的数量似乎并未消散。提高认识运动成功地提醒公众注意这一特定犯罪；然而，很难知道它们是否减少了潜在的受害者人数(特别是考虑到许多受害者在成为受害者之前就知道了这一罪行；见Whitty，2013年出版)。对这一特定罪行的起诉是非常耗费资源的，其对减少犯罪的影响尚不清楚。我们选择了在不同领域拥有专长的合作伙伴(国内和国际)，包括：执法、情报、第三部门和工业。他们有不同的知识可以分享，也可能使用不同的方法来解决这个问题(例如，行业可以筛选和发现欺诈者，执法部门可以追踪罪犯并提高认识，第三部门可以实施方法来保护公民并使他们更具弹性)。从学术的角度来看，多学科方法增加了我们发现和预防这种犯罪的机会。了解易受影响的人的类型、使人更容易受到伤害的情景条件以及用于说服目标相信互动是真实的并说服他们将钱交给欺诈者的方法和材料(例如，在线个人资料、消息、通信方法)，对于制定打击这一问题的方法至关重要。将这些知识与更多的技术知识结合起来，为我们提供了更强的检测和预防能力。例如，技术指标，如电话号码、IP地址、与被盗身份材料的链接、说服性信息的风格模式，提供了对犯罪的更丰富的了解，并提供了更多的变量来帮助侦查。此外，在检测MMF时开发的任何工具都需要让最终用户相信他们被诈骗的可能性(当罪犯试图说服受害者相信不同的情况时，这一点尤其困难)。鉴于此，在开发有用的方法和信息以说服潜在受害者他们正在与罪犯互动时，人机界面方法是至关重要的。最后，我们需要考虑我们可能用来检测和预防MMF的个人数据类型的伦理问题。

Fraud can be broadly defined as trickery used to gain a dishonest advantage, usually financial, over another person or organisation. Mass-marketing fraud (MMF) is a type of fraud that exploits mass communication techniques (e.g., email, Instant Messenger, bulk mailing, social networking sites, telemarketing) to con people out of money. It is a scam that targets victims in most countries. The 419 or 'Nigerian' scam is one example. In this scam the fraudster requests upfront fees with the promise that the victim will recover large sums of money in return for little effort. However, not all mass-marketing frauds con the victim with the promise of making large sums of money. In the romance scam, for example, the criminal pretends to develop a romantic relationship with the victims and later requests money to help them, especially in a crisis. In the charity scam, victims believe they are giving money to a genuine charity.The proposed project will develop novel techniques to detect and prevent MMF. Through its multi-disciplinary approach and close focus on co-designing the solutions with its range of project partners and testing them in-the-wild during live MMF-detection settings, the project will lead not only to new scientific understanding of the anatomy of MMF but also to tools and techniques that can form the basis of practical interventions in tackling such fraud. Working with partners outside of academic is crucial to the success of this project. Over the years various types of organisations have worked hard to detect and prevent MMF (often in silos) - with some methods appearing to be somewhat effective. Nonetheless, the numbers of victims do not appear to be dissipating. Awareness campaigns have succeeded in alerting the public to this particular crime; however, it is difficult to know if they have reduced the potential number of victims (especially, given that many victims are aware of the crime prior to becoming victims; see Whitty, 2013, in press). Prosecution for this particular crime is very resource intensive, and its effects on crime reduction are unknown. We have chosen partners (national and international) who have specialities in different fields including: law enforcement, intelligence, third sector, and industry. They have different knowledge to share and also can potentially tackle the problem using different methods (e.g., industry can screen out and detect fraudsters, law enforcement can trace criminals and raise awareness, third sector can implement methods to protect citizens and to make them more resilient).From an academic perspective, a multi-disciplinary approach increases our chances of detection and prevention of this crime. Understanding the types of people susceptible, the situational conditions that make a person more vulnerable, and the methods and materials (e.g., online profiles, messages, communication methods) used to convince the target that the interaction is authentic and persuade them into giving up their money to a fraudster is crucial in the development of methods to combat this problem. Combining this knowledge with more technical knowledge provides us with a much greater capability to detect and prevent. For example, technical indicators, such as phone numbers, IP addresses, links to stolen identity material, stylistic patterns of persuasive messaging provide a much richer understanding of the crime and provides a greater number of variables to assist in detection. In addition, any tool developed in detection of MMF needs to convince the end-user of the likelihood that they are being scammed (which is especially difficult when the criminal is attempting to persuade the victim to believe otherwise). Given this an HCI approach is crucial when developing usable approaches and messages that persuade the potential victim that they are interacting with a criminal. Finally, we need to consider the ethics of the type of personal data we might utilise to detect and prevent MMF.