In response to the reported increase in accounting crime and suspected fraud in conjunction with the adjustment subsidies initialized during the covid-19 pandemic of 2020, we conducted a study examining accounting manipulation. Based on an explanatory theory for fraud, the Fraud Triangle, we theorize that the financial circumstances that companies experienced during the pandemic and the government subsidies created an environment where accounting manipulation could occur. We obtained a unique set of unpublished data from the Swedish Tax Agency. Two common models for detecting accounting manipulation were then applied to a large sample of Swedish companies. We used a version of the Jones Model to detect potential earnings management and Benford’s Law to detect fraudulent manipulation of the firms’ reported loss of revenue. Our results indicate earnings management and fraudulent reporting for some industries but no broad indication of systematic accounting manipulation across all industries. However, we suggest future research on this topic to further understand how accounting manipulation occurs in distressed industries.

As popularity of online banking services has grown exceedingly among Finnishconsumers, it has become necessary for banks to provide their customers withsafety advice against the emerging threat of card-not-present fraud (CNP) in orderto protect them from monetary losses. However, it is unclear how effective thisadvice is and how well it is filling its purpose. This study aims to fill this gap andexamines the advice provided by Finnish banks in order to protect their customersfrom CNP-frauds by applying an economical approach to the criminological field.A multilayered approach including a literature review, a web page qualityassessment and a qualitative interview was used for this purpose. Contributing toexisting literature on the enabling and constraining influences within the financialindustry, this study increases the understanding of why Finnish banks’ arehomogenously tilting the balance towards their own private interests rather thanpublic good.

The chargeback process is costly for the merchant. Not only does the merchant lose the revenue from the purchase, but it must also pay handling fees to the bank and risks never getting paid for provided service. The purpose of this study is to examine and investigate how to prognosticate future chargebacks by using machine learning in form of Artificial Neural Network on transaction data. Doing so can be used to minimize and decrease financial costs for the merchant. The study indicates that it’s complex to prognosticate chargebacks, but illuminates that it’s possible under certain circumstances. The created model has been concluded to be more suitable as a compliment, rather than a substitute for the current Rule-Based classification system. The model should be implemented based on economic analysis since it can be used to reduce costs and contribute to profitability over time. Furthermore, the study highlight lessons learned and complementary research areas for future studies. / Återbetalningar medför stora kostnader för handlare i form av förlorade intäkter vid återbetalning av transaktionssumman, samt tillkommande handläggningsavgifter i processen. Syftet med rapporten är att utvärdera och undersöka möjligheten att prognostisera framtida återbetalningar, genom att applicera maskininlärning i form av Artificiellt Neuralt Nätverk på transaktionsdata. På så sätt kan återbetalningar minimeras och reducera finansiella kostnader hos handlaren. Studien påvisar att det är komplext att predicera återbetalningar, men att det är möjligt under särskilda omständigheter. Modellen som skapats har konstaterats mer lämpad som ett komplement till det aktuella regelbaserade klassificeringssystemet än ett substitut. Utifrån en ekonomisk analys klargörs att algoritmen bör implementeras för att reducera kostnader och på sikt bidra till lönsamhet. Studien belyser även lärdomar, samt kompletterande forskningsområden för framtida studier.

This study aims to create a Machine Learning model and investigate its performance of detecting fraudulent user behaviour on an e-commerce platform. The user data was analysed to identify and extract critical features distinguishing regular users from fraudulent users. Two different types of user data were used; Event Data and Screen Data, spanning over four weeks. A Principal Component Analysis (PCA) was applied to the Screen Data to reduce its dimensionality. Feature Engineering was conducted on both Event Data and Screen Data. A Random Forest model, a supervised ensemble method, was used for classification. The data was imbalanced due to a significant difference in number of frauds compared to regular users. Therefore, two different balancing methods were used: Oversampling (SMOTE) and changing the Probability Threshold (PT) for the classification model.  The best result was achieved with the resampled data where the threshold was set to 0,4. The result of this model was a prediction of 80,88% of actual frauds being predicted as such, while 0,73% of the regular users were falsely predicted as frauds. While this result was promising, questions are raised regarding the validity since there is a possibility that the model was over-fitted on the data set. An indication of this was that the result was significantly less accurate without resampling. However, the overall conclusion from the result was that this study shows an indication that it is possible to distinguish frauds from regular users, with or without resampling. For future research, it would be interesting to see data over a more extended period of time and train the model on real-time data to counter changes in fraudulent behaviour.

<p>Consumer behaviour has and may increasingly have a vital role to play in protecting personal data. Understanding the behaviours of consumers in preventing identity theft and detecting identity fraud is therefore key to creating programs that minimize exposure and potential loss. In this study, based on the Theory of Planned Behaviour (TPB) and Protection Motivation Theory (PMT), an exploratory study elicited salient beliefs about identity theft prevention and detection behaviours. These beliefs were then used to create a survey to measure the strength of the salient beliefs, attitudes, intentions and behaviours, which was administered online and produced 351 valid responses. Statistical analysis was performed on eight behavioural groups, based primarily on principal component analysis of twelve behaviours. The groups were: using physical security, practicing password security, monitoring bank accounts and credit cards, getting a credit report, checking the land registry, using 'remember my password', clicking on a link in an e-mail, and giving out personal information over the phone. Results showed that beliefs with a significant influence on consumer intentions for a given behavioural group were a mix of beliefs about identity theft in general and beliefs about the behaviours in that group. While attitudes towards behaviours of consumers in any specific group had a significant influence on the intent to perform behaviours peculiar to that group, they had virtually no impact on the intent to perform behaviours in other groups. The intent to perform identity theft prevention and identity fraud detection behaviours uniformly had a statistically significant influence on actual reported behaviour, but much of the variance in behaviour was unexplained. An analysis of qualitative responses showed that gender, language and age all had significant impacts on respondents' likelihood of mentioning specific vulnerabilities, and prevention and detection measures.</p> / Doctor of Philosophy (PhD)

No / The degree of insecurity occasioned by fraudulent practices in Nigeria has been of great concern economically, especially as it relates to overseas transactions. This paper was designed to mitigate this problem for Nigeria and countries with similar dispositions. Based on a survey involving field trip to Nigeria, the paper examines the general security situation in Nigeria and its mutual impacts with computerisation, miniaturisation and Location-Based Authentication (LBA). It was discovered that both computerisation and miniaturisation had some negative effects on cybersecurity, as these were being exploited by fraudsters, especially using ‘advance fee fraud;’ popularly called 419. As a countermeasure, the research examined the possibility of using LBA and further digitisation of the GSM Mobile country codes down to City/Area codes along with GSM Mobile/Global Positioning System (GPS) authentications. Where necessary, these could be combined with the use of a web-based Secret Sharing Scheme for services with very high security demands. The anticipated challenges were also examined and considered to be of negligible impacts; especially roaming. / Petroleum Technology Development Fund (PTDF)

Yes / The state of insecurity occasioned by fraudulent practices in Africa has been of concern economically, both at home and abroad. In this paper, we propose ways to mitigate this problem, using Nigeria as a case study. Based on surveys in West Africa, the paper examines the security situation in the continent and its mutual impacts with computerisation, miniaturisation and Location-Based Authentication (LBA). It was discovered that computerisation and miniaturisation had negative effects on cyber-security, as these were being exploited by fraudsters, using advance fee fraud; called 419. As a countermeasure, the paper examines the possibility of using LBA and digitisation of the GSM Mobile country codes down to city/area codes along with GSM/GPS authentications. These could also be combined with the use of a web-based Secret Sharing Scheme for services with very high security demands. The challenges of roaming were also examined and considered to be of negligible impact. / Petroleum Technology Development Fund (PTDF)

Den tidigare forskningen har visat att sexuella övergrepp ofta medför svåra konsekvenser för offret. Vidare finns en risk för ytterligare trauma i form av sekundär viktimisering, om offret möts av skuldbeläggning från sin privata omgivning eller rättsväsendet. I denna studie studerades svenska tingsrättsdomar för att undersöka skuldbeläggning av målsäganden i våldtäktsdomar, samt för att jämföra eventuella skillnader i detta avseende med misshandel- eller bedrägerimål. Vidare undersökte studien huruvida det förekommer utmanande av skuldbeläggning i domarna. Studien har utgjorts av en kvalitativ dokumentstudie med en tematisk innehållsanalys. Vidare har studiens teoretiska ramverk utgjorts av begreppen våldtäktsmyter, skuld och tron på en rättvis värld.  Fynden visade att våldtäktsmyter kom till uttryck i majoriteten av våldtäktsdomarna, vilket i många fall gav upphov till skuldbeläggning av målsäganden. Oberoende av måltyp identifierades även olika typer av skuldbeläggande diskussioner. I våldtäktsdomarna handlade diskussionerna främst om huruvida målsägande genom sitt agerande gett uttryck för samtycke. Sett till misshandelsdomarna handlade det istället om diskussioner kring målsägandens eventuella provokation och tilltalades rätt till nödvärn. Gällande bedrägeridomarna identifierades diskussioner kring hur målsäganden hade kunnat undvika bedrägeriet. Slutligen identifierades utmanande av skuldbeläggning av målsäganden i domarna, främst i våldtäktsdomarna. Avslutningsvis visar studien att andelen domar där skuldbeläggning förekom inte skiljde sig avsevärt mellan de olika måltyperna. Däremot identifierades ett större antal fynd av skuldbeläggning i våldtäktsdomarna än i domarna för de andra måltyperna. Således tycks skuldbeläggning av målsägande i våldtäktsmål vara vanligare, sett till mängden fynd, jämfört med skuldbeläggning av målsägande i misshandel- och bedrägerimål.

This study introduces industry competition factors to fraud models to examine how competition associates with fraud risk. I argue that industry competition eclipses many firm-level determinants in their association with fraud risk, and that the cost of poor information elevates fraud risk as competition increases. I find that fraud risk is higher for firms in industries with 1) more substitutable products and services, 2) greater threats of new entry, and 3) larger incumbent pools of competitors, and that substitution exceeds every firm-level variable except size in its relevance with fraud risk. Cross-sectionally, I provide evidence that industry-wide non-adoption of advanced cost analytics (i.e. using obsolete, distortionary standard costing practices) may exacerbate the fraud-risk effects of competition, especially product substitution: a one standard deviation increase in substitution associates with over double the fraud risk for firms in industries typified by obsolete costing practices. I also find that different strategies vary in their fraud-offsetting associations dependent on the type of competition most prevalent in an industry. Together, these findings shed light on how the effects of industry competition may subsume or surpass most firm-level fraud determinants and provide evidence of previously unidentified drawbacks of obsolete cost accounting systems. / Doctor of Philosophy / Elements of industry competition help explain a firm's fraud risk. I find that bringing competition variables into firm-level fraud models helps explain a large portion of the firm's fraud risk, and that the effects of competition more strongly associate with fraud risk than most firm-level attributes. The results also indicate that the effects of competition on fraud risk may be even worse in industries where obsolete cost accounting practices remain widespread: the effects of price competition in such industries associates with significantly greater fraud risk than in other industries. Additional findings include the implied fraud-risk-reducing effects of different business strategies, depending on which type of competition is most intensive around a firm. Altogether, this study sheds light on the importance of including industry competition effects when assessing fraud risk, especially when a firm's or its peers' cost accounting system quality is poor and price competition is high.

According to security.org, as of 2023, 65% of credit card (CC) users in the US have been subjected to fraud at some point in their lives, which equates to about 151 million Americans. The proliferation of advanced machine learning (ML) algorithms has also contributed to detecting credit card fraud (CCF). However, using a single or static ML-based defense model against a constantly evolving adversary takes its structural advantage, which enables the adversary to reverse engineer the defense's strategy over the rounds of an iterated game. This paper proposes an adaptive moving target defense (MTD) approach based on deep reinforcement learning (DRL), termed RESONANT to identify the optimal switching points to another ML classifier for credit card fraud detection. It identifies optimal moments to strategically switch between different ML-based defense models (i.e., classifiers) to invalidate any adversarial progress and always stay a step ahead of the adversary. We take this approach in an iterated game theoretic manner where the adversary and defender take turns to take their action in the CCF detection contexts. Via extensive simulation experiments, we investigate the performance of our proposed RESONANT against that of the existing state-of-the-art counterparts in terms of the mean and variance of detection accuracy and attack success ratio to measure the defensive performance. Our results demonstrate the superiority of RESONANT over other counterparts, including static and naïve ML and MTD selecting a defense model at random (i.e., Random-MTD). Via extensive simulation experiments, our results show that our proposed RESONANT can outperform the existing counterparts up to two times better performance in detection accuracy using AUC (i.e., Area Under the Curve of the Receiver Operating Characteristic (ROC) curve) and system security against attacks using attack success ratio (ASR). / Master of Science / According to security.org, as of 2023, 65% of credit card (CC) users in the US have been subjected to fraud at some point in their lives, which equates to about 151 million Americans. The proliferation of advanced machine learning (ML) algorithms has also contributed to detecting credit card fraud (CCF). However, using a single or static ML-based defense model against a constantly evolving adversary takes its structural advantage, which enables the adversary to reverse engineer the defense's strategy over the rounds of an iterated game. This paper proposes an adaptive defense approach based on artificial intelligence (AI), termed RESONANT, to identify the optimal switching points to another ML classifiers for credit card fraud detection. It identifies optimal moments to strategically switch between different ML-based defense models (i.e., classifiers) to invalidate any adversarial progress and always stay a step ahead of the adversary. We take this approach in an iterated game theoretic manner where the adversary and defender take turns to take their action in the CCF detection contexts. Via extensive simulation experiments, we investigate the performance of our proposed RESONANT against that of the existing state-of-the-art counterparts in terms of the mean and variance of detection accuracy and attack success ratio to measure the defensive performance. Our results demonstrate the superiority of RESONANT over other counterparts, showing that our proposed RESONANT can outperform the existing counterparts by up to two times better performance in detection accuracy and system security against attacks.