The University of Iowa, an educational institution with a considerable reputation for its engineering and technology research departments, recently released a paper on detecting fraud for online businesses using blockchain technology. Researchers at the university looked into the effectiveness of blockchain technology in detecting objective fraud and ultimately, its applicability in the online fraud security market.
Blockchain technology, can be used to secure user data, ratings and other valuable information on online platforms and businesses such as marketplaces; its immutable, transparent and robust infrastructure enables data to be stored in an unalterable ecosystem, eliminating the possibility of information and data manipulation.
Blockchain’s Weakness in Rating or Subjective Fraud
The term rating or subjective fraud in the online business or marketplace industry stands for user ratings or consumer feedback which future consumers can rely on to make a purchase from online merchants. Marketplaces like Amazon, eBay and Fiverr have built-in rating systems for customers to rate the service or product they have received.
“User-driven rating systems (i.e., eBay or Amazon) compute their rating scores based on users’ rating. In user-driven rating systems, the rating score can be calculated either as the difference between all positive and negative scores. or as the average of all ratings (e.g., Amazon),” said the research team.
Prior to conducting their study, researchers at the University of Iowa believed that blockchain technology will flawlessly execute most of the operations of an online platform when it comes to data management and storage. However, the team discovered one critical weakness of blockchain technology, which ironically is its most desirable strength; immutability.
Ratings are an important aspect of online platforms as they determine the legitimacy of a merchant or a seller. However, many merchants tend to rate themselves through third-party service providers to make themselves stand out on the platform.
For instance, there are service providers that offer a certain number of ratings and comments on a marketplace for cash or digital payment in return. In private platforms or ledgers, network administrators can overturn these fraudulent votes or ratings. The so-called “rating manipulation” can be avoided with reports sent to the network administrator.
With blockchain technology, it is virtually impossible to overturn these ratings as data and information stored on the blockchain network is irrefutable.
To accurately measure the efficiency of blockchain technology in online fraud detection, the research team looked into the applicability of the blockchain in the following types of attacks:
- Constant attack: fraudulent reviewer consistently provides unfair ratings.
- Camouflage attack: a strategical fraudulent reviewer injects fair ratings to camouflage himself/herself. This makes fraud detection incredibly difficult.
- Whitewashing attack: a fraudulent reviewer injects unfair ratings to a target for a certain period. The reviewer washes his reputation by creating a new account.
After testing blockchain technology by applying it to the three situational attacks, the team at Iowa University discovered that the blockchain is effective in preventing objective fraud, in which all cases of fraud or mismanagement of data is based on factual information such as numbers.
However, in subjective fraud, which occurs in most online businesses, marketplaces, and platforms, blockchain technology is less efficient as there is no ground for the technology to base on to detect fraud.
“Blockchain systems are effective in preventing bad mouthing and whitewashing attack, but they are limited in detecting ballot-stuffing under Sybil attack, constant attacks and camouflage attack,” said the team.
Based on the findings of the research team, online businesses must acknowledge that the blockchain is not always useful and applicable when it comes to data. Companies need to be cautious when implementing blockchain technology to manage objective and subjective data, as blockchain-based systems could have issues dealing with subjective information.