E-mail Print PDF

The unprecedented rise of e-commerce during the last years has massively increased online piracy and counterfeiting, especially on auction platforms and online market places on account of the large amount of retailers, whose products are never validated on authenticity before they are distributed. Even traditional online shops and B2B platforms are nowadays in the focus of counterfeiters; retailers may deliberately or unintentionally sell products to customers. Concerned customers will obtain a low-quality product, which in some cases may be even dangerous to life. Besides the customers, online platforms and enterprises suffer much damage because of image loss and financial losses. Due to the huge anonymity of the Internet, it is practically impossible to pointedly reveal counterfeiters and to persistently thwart their unscrupulous business.

The WDI-Lab thus aims to develop semi-automatic solutions in order to discover fake articles in the Internet. Seizing on our technologies successfully implemented in the past, such as query generation and product matching, new strategies with respect to counterfeit detection are now elaborated, like product clustering and price scoring. With this, a list of products can be iterated and likely imitations be discovered. Even though this approach may be not sufficient to take legal actions against counterfeiters, it will be a great support for any owner of a website to suspend or remove suspicious offers resp. retailers. Especially when it comes to large online platforms, a manual approach seems completely inconceivable so that semi-automatical solutions become inevitable.

Our recently developed prototype is already capable to extract different kind of products from different categories and different data sources (like eBay, Amazon, Mercateo, Goolge Product Search etc.) that are to be analyzed in order to detect counterfeit candidates. We primarily focus on frequently imitated products, such as cosmetics, luxury goods, apparel and medication, although our approach can be used in almost any domain. We are looking forward to cooperate with any interested industrial partner dealing with counterfeit detection, and at the same time aim to seize on the knowledge and experience of partners, in order to adjust our software in the best possible way.

Last Updated on Thursday, 18 October 2012 11:57  

Logo BMBF Logo Unternehmen Region