Online shopping has become commonplace. Due to the popularity, also the market for so-called "fake shops" – i.e. ripping of customers with fraudulent and fake portals – has been ever growing since and is one of the largest threats for Austrian customers according to cybercrime reports. One of the most prominent examples in the media was the damage of 440k euros caused by a single Munich native who offered about 750 popular products such as coffee machines, mobile phones, game consoles in 21 different online shops - but none of the products ever got delivered.
Besides counterfeit sales platforms with trademark infringed goods, one of the primarily target for fake-shops include offerings of digital goods (e.g. content streaming, navigation software, etc.). Exposing such fake offerings is an extremely difficult task especially for under-savvy online consumers (such as senior citizens). Since it is almost impossible to track down the fraudsters much less to enforce consumer rights in such situations, it is important to take preventive measures.
The Internet Ombudsstelle was established as e-commerce service in July 2013. On its platform Austrian consumers receive regular updates, warnings and tips how to protect themselves against online fraud. In addition, a blacklist of fraudulent online shops is maintained. One hundred and fifty online shops are submitted every week for manual inspection. One of the biggest problems with fake-shops is that, as soon as they are exposed, they immediately disappear from the internet and are put back online in a slightly modified form and under new domains. Often, dozens or more of these copies exist online at the same time. Therefore, it's important to act quickly!
The goal of the KOSOH project is to capture the intrinsic knowledge of ÖIAT employees in their abilities of identifying fake-shops and to provide technical methods for assisting and accelerating their work. This includes applying AITs expertise in machine learning (neural networks) technologies for automatically classifying web pages based on their structural information (such as CSS, DOM, Javascript, referenced media, etc.), as well as using image similarity tools for identifying kite marks. The model of fake shop measures which are built up and trained in KOSOH allow providing similarity and risk assessments scores for any online-shop.
KOSOH is an important preventive measure in the detection of fake-shops and will be available as browser plugin for Austrian consumers. A central component of the project is the final evaluation of existing counter measures. The report takes into account existing protection mechanisms, their shortcomings the findings, quality and reliability of the given KOSOH approach and concludes with recommendations for policy makers.
The project is funded by the KIRAS security research programme of the Federal Ministry of Finance.