Another Way of Doing Fraud Analysis in Elections
As widely reported, presidential candidate number three Ganjar Pranowo asked his supporting parties, including the PDIP, in the House of Representatives (DPR) to investigate alleged election fraud.
Ganjar stressed that alleged fraud practices during the election must be seriously addressed. “If the DPR is not ready with the right of inquiry, I encourage the use of the right of interpellation or working meeting,” said the former governor of Central Java at his residence in South Jakarta on Friday, February 16.
During a meeting with his campaign team on Feb. 15, Ganjar presented thousands of messages in the form of photos, documents, and videos of various alleged election frauds.
Meanwhile, presidential candidate number one Anies Baswedan said that his three supporting parties in the Coalition for Change were ready to use the right of inquiry in the DPR to investigate the election fraud. The three parties are NasDem, PKB, and PKS.
Anies made the statement in response to Ganjar’s initiative, which he deemed a good move. The former governor of Jakarta expressed confidence that his coalition would support the PDIP, which is the largest faction in the DPR to investigate allegations of fraud through legislative channels.
Anies also mentioned that the Coalition for Change is ready to provide pieces of evidence to support the inquiry process.
The fraud will be did, because the personal has a motivation to get a achievement. It gains or benefits for the fraudster, as emphasized by the definition of fraud provided by the Oxford Dictionary.
The Oxford Dictionary defines fraud as follows: Wrongful or criminal deception intended to result in financial or personal gain.
In any business, all activity can be fraudulent. People have responsibility for doing their task. In a simple example, the dustman collects some garbage in a sector area. There are 120 trash bins, and in normal conditions, the dustman can collect them in 1 hour. In a day, a rock concert has shown up in this area. There is much garbage in their field, and the dustman need an hour to collect all of it. So the dustman must do extra work in an hour. The dustman concluded that they must collect all of the garbage in the field and all of the resident house in 1 hour because of the cost. So the dustman just collects the garbage in half of the area (60 trash bins in the resident house). So all of the garbage in that area has not been collected.
Based on that cause, it concluded in a hypothesis formulated by Donald R. Cressey in his 1953 book Other People’s Money: A Study of the Social Psychology of Embezzlement:
Trusted persons become trust violators when they conceive of themselves as having a financial problem which is non-shareable, are aware this problem can be secretly resolved by violation of the position of financial trust, and are able to apply to their own conduct in that situation
verbalizations which enable them to adjust their conceptions of themselves as trusted persons with their conceptions of themselves as users of the entrusted funds or property.
Based on that hypothesis, it can be formulated in fraud triangle.
- Pressure, An individual will commit fraud because a problem is experienced of financial, social, or any other nature.
- Opportunity, The second leg of the model, and concerns the precondition for an individual to be able to commit fraud.
- Rationalization, The psychological mechanism that explains why fraudsters do not refrain from committing fraud and think of their conduct as acceptable.
Because of the pressure, the fraudster can relate the real condition to the fake condition using rationalization for anything to be acceptable. Fraudulent activities can only be committed when the opportunity exists for the individual to resolve or relieve the experienced pressure or problem in an unauthorized, concealed, or hidden manner.
Let’s go back to the Dustman tale. Because of cost management, the dustman has pressure to reduce their costs. They are using the opportunity to collect all of the garbage at midnight, when all of the residents are asleep. So the dustman can collect the garbage as they want; not all the residents get to collect the garbage because the residents are asleep and do not see how the dustman works. They are doing that based on rationalization: because the rock concert has ended after midnight, the dustman will collect the garbage at midnight. It’s a perfect fraud.
Based on how the fraud is coming, some procedures must be implemented to detect the fraud and create a fraud prevention program. It will be automated in a cycle to identify the fraud. A cycle of fraud prevention is needed.
That figure introduces the fraud cycle, and depicts four essential activities:
- Fraud detection: Applying detection models on new, unseen observations and assigning a fraud risk to every observation.
- Fraud investigation: A human expert is often required to investigate suspicious, flagged cases given the involved subtlety and complexity.
- Fraud confirmation: Determining true fraud label, possibly involving field research.
- Fraud prevention: Preventing fraud to be committed in the future. This might even result in detecting fraud even before the fraudster knows they will commit fraud.
Between fraud prevention and fraud detection, there is fraud investigation and fraud confirmation, which are summarized based on the upcoming fraud. The fraud anomaly will be detected based on an uncommon plot that is out of the normal pattern. The uncommon plot will be investigated using modeling based on learning (supervised or unsupervised learning), and it will be summarized in the evaluation for fraud confirmation. Based on the results, a statement can be used to make a decision about fraud prevention.
Data
Because of business needs, the data is a good commodity to make anything good decision. The business leader makes a decision based on reality. The row data are containing activity log based on fact. It based on the previous article to describe about data https://medium.com/@adi_pradana14/data-956d4e08e689 .
The coming fraud can be identify based on the data what stored in a database or dataset. The pattern will be shown based on the common data, so the uncommon things will be shown too.
Fraud is an uncommon, well-considered, imperceptibly concealed, time-evolving and often carefully organized crime which appears in many types of forms. — Van Vlasselaer et al. (2015).
Because of Uncommon things, the “real-path” will be shown in right way and make a difference with the fraud, so the fraud will be shown in another way, left way. In large datasets, it can be mined to discover interesting patterns, models, and knowledge; this is called data mining.
Literally, the fraud will be identified using predictive and descriptive analytics. In predictive, the observer will identify the fraud like a football pundit who does match analysis before a football match based on match history. In real-life conditions, the observer must create a flag for each data point on the dataset. Utilize supervised learning using that flag and start to choose regression or classification modeling to create a prediction based on historical data in the dataset.
Meanwhile, in descriptive analytics, the observer will identify the fraud like a postman giving mail in a clustered house. The postman will deliver the mail based on the house number, area, and postcode. In the real world, the observer must divide the data based on the cluster, use unsupervised learning, and choose clustering modeling to make a description based on historical data in the dataset.
It concluded that the fraud will be identified using unsupervised learning (clustering) and supervised learning (classification and regression).
Implementation of the fraud analysis in US election 2020 using python will be posted in this article : https://medium.com/@adi_pradana14/d-i-y-fraud-analysis-in-election-e27971c76191
Source:
https://en.tempo.co/read/1835881/jokowi-indifferent-to-2024-election-fraud-inquiry
Baesens, B., Van Vlasselaer, V., & Verbeke, W. (2015). Fraud analytics using descriptive, predictive, and social network techniques: A Guide to Data Science for Fraud Detection. John Wiley & Sons.
Han, J., Pei, J., & Tong, H. (2022). Data mining: Concepts and Techniques. Morgan Kaufmann.
