EUR
en
- es
- pt
- fr
- de
- it
Published on 2023-11-26
Amazon uses Artificial Intelligence (AI) tools, including machine learning models, advanced linguistic models (LLM), and graph neural networks, to analyze, detect, and eliminate fake reviews, as well as to assure users that the reviews published are reliable.
The e-commerce company believes that a fundamental part of the shopping experience on its platform is making it easy for customers to share their opinions on products simply, thus helping other customers make informed purchasing decisions.
However, among all the reviews published on the platform, there are also those that are fake, shared by some users to 'take advantage' of the shopping experience on Amazon. These reviews intentionally deceive customers by providing information that is not impartial. Techniques such as review hijacking are even used to sell counterfeit products at competitive prices.
In this regard, to put an end to this type of fraudulent publications, the company led by Jeff Bezos uses a series of AI-powered tools to identify and eliminate fake reviews and to ascertain which ones are genuinely authentic.
As explained in a post on its blog, before publishing a review, Amazon analyzes this information looking for risk indicators that help detect if it might be fake. Thus, if signs are found, the company acts quickly to 'block or remove it' and, if necessary, 'take additional measures'.
For example, if a review is determined to be fake, measures such as revoking the 'privilege' of posting reviews, blocking the accounts of the offenders, or even pursuing legal actions are taken.
On the other hand, if an opinion is suspicious but cannot be clearly confirmed as a fake review, Amazon uses specially trained researchers and experts to 'identify abusive behaviors'. Specifically, this staff is responsible for 'searching and reviewing other risk signals before making a decision'.
Likewise, Amazon also uses the 'latest advancements' in AI to detect these fake reviews, as well as manipulated ratings, fraudulent customer accounts, and other abuses 'before customers see them'.
As detailed by the company, they use machine learning models to analyze a multitude of data. In this case, considerations such as whether the seller has invested in advertising, which allows them to generate an additional number of reviews, are taken into account. Also, aspects such as the submission of abuse reports regarding that seller or the review history are analyzed.
Following this line, Amazon also uses large language models, along with other natural language processing techniques, to 'analyze anomalies' that may indicate that a review is fake or even incentivized by a reward such as a gift card or a free product in exchange.
Another tool used by the platform is graph neural networks (GNN), with which it can understand complex relationships and behavior patterns. As detailed, this is a 'crucial' technology because, to make a correct assessment of a review, it is necessary to consider both the information related to the type of customer who writes it and the type of product.
Thanks to the combination of these AI technologies, Amazon ensures it can identify fake reviews with 'greater accuracy' because it involves a more in-depth analysis than that achieved with superficial abuse indicators. In fact, the company emphasized that with AI, it is possible to 'identify deeper relationships between potential offending behaviors'.
'Maintaining a reliable and secure shopping experience is our top priority,' said Amazon's Head of External Relations and Trusted Reviews, Rebecca Mond, while asserting that they will continue to develop new ways to prevent the publication of fake reviews and protect customers to 'shop with total confidence'.
For his part, the head of Amazon's Fraud and Abuse Prevention team, Josh Meek, commented that preventing fake reviews is important because, not only 'do millions of customers trust the authenticity of the reviews to make purchasing decisions,' but 'also millions of brands and business partners' trust Amazon to 'accurately identify reviews that are fake and prevent them from reaching customers'.
Finally, Amazon highlighted that, during the year 2022, it managed to proactively identify and block over 200 million reviews suspected of being fake globally.
COMMENTS
No customer comments for the moment.