Startup Launches 'Fake-Detecting AI' to Foster a Healthy IT Ecosystem

IT Startup Leads Creation of Healthy Online Ecosystem Using Artificial Intelligence (AI)
Actively Utilizing AI to Enhance Service Convenience and Strengthen User Protection

Startups that detect fake and false listings and reviews based on artificial intelligence technology and work to create a healthy IT ecosystem are gaining attention. Graphic = gettyimage

[Asia Economy Reporter Kim Heeyoon] Amid the surge in contactless consumption due to the impact of the novel coronavirus infection (COVID-19), the technology development by IT companies to detect fake and false listings and reviews is becoming a hot topic.

According to industry sources on the 19th, IT companies utilizing artificial intelligence (AI) technology to wage the so-called ‘war against fakes’ have been confirmed to be protecting users by fostering a healthy ecosystem and market.

Danggeun Market Develops Its Own AI Machine Learning Automation Technology

First, Danggeun Market, whose usage has surged so much that the new phrase ‘Danggeunhaseyo?’ has become popular recently, applies its self-developed AI machine learning technology to filter products. The company explained that it has actively utilized AI technology from the early stages of the service to create a trustworthy transaction environment. AI performs real-time filtering to pre-screen posts involving prohibited items such as alcoholic beverages, tobacco, animals, and counterfeit goods, and automatically recognizes and handles problematic posts.

Danggeun Market fundamentally blocks the activities of professional sellers, known as ‘kkun.’ Activities by merchants aiming for profit beyond connecting local residents are completely restricted. If a post is recognized as belonging to a professional seller through real-time AI screening, the post is automatically set to private, and the user’s activities are also restricted. In addition, the company’s representative explained that AI technology is strongly responding to inappropriate posts that harm service usability, harmful content, false or advertising posts, and fraudulent activities.

MarkVision Detects Counterfeit Products Using AI Technology

AI company MarkVision provides an automated service that handles everything from determining intellectual property (IP) infringement to managing, reporting, and deleting counterfeit product listings.

MarkVision’s self-developed AI algorithm has been trained on millions of ‘image’ and ‘text’ data. Using deep learning image recognition technology, it quickly and accurately identifies visually similar products among millions of items sold on e-commerce platforms. Furthermore, it detects counterfeit products by analyzing data such as product descriptions, prices, and purchase reviews of similar products through machine learning.

In fact, a global beauty brand had been detecting about 300 counterfeit products per month with its own staff, but after applying the MarkVision platform, it reported 3,000 counterfeit products in just one month, proving its performance, according to a company representative.

MarkVision monitors in real-time whether products distributed on global e-commerce sites such as Taobao, Tmall, AliExpress, eBay, and Amazon infringe on intellectual property rights based on its AI algorithm. Company representatives can review the counterfeit product list collected by AI and report it to each e-commerce site with a single click. A MarkVision representative emphasized, “From a company’s perspective, it can save enormous human resources and costs and also prevent consumer damage.”

‘Featuring’ Filters Fake Followers to Measure Influencers’ Actual Impact

Featuring, an AI-based SNS influencer analysis solution company, provides services that measure and analyze influencers’ actual impact. It analyzes various data such as followers’ tendencies, images, and comments using 40 different indicators to select influencers suitable for specific brands and products.

Featuring developed a system that removes fake followers, the most important task in measuring an influencer’s actual impact. The company explained that by filtering out false followers using its self-developed AI-based data engine, objective impact measurement is possible.

First, it builds pattern sets to filter fake likes and applies about 150 patterns to detect macro or ghost account activities, distinguishing fake accounts and likes. Additionally, it technically measures the suitability between products and audiences to provide objective data. A Featuring representative explained, “We provide sophisticated analysis services so that companies preparing influencer advertisements can predict the best results with a more efficient budget.”