Issue 55/2019

The CEA states that the primary version is the paper one.

Łukasz Lipiński, Michał Bernardelli

Detekcja wyłudzeń klików przez boty internetowe

***

Click fraud detection rules

Abstract
Effective detection of clicks on websites done by automatic computer programs is a valuable tool in the fight against this type of fraud and gives immediate measurable benefit in the form of savings for poorly targeted advertising. The purpose of the study described in the article was to present the unsupervised learning method, which results in a set of rules that allow for effective detection of bots that are responsible for the click frauds, assuming that their behavior differs in some aspects from human behavior. This analysis proves that involving device type as an extra variable improves the effectiveness of rules used for fraud detection and that the proposed algorithm provides a flexible and efficient solution for the given problem.

Article: PDF

Table of contents of issue 55

Copyright © Kolegium Analiz Ekonomicznych Szkoły Głównej Handlowej w Warszawie 2011-2024   ISSN 1232-4671
sie wyburaczylo