An Analysis Model For Detecting Click Farm in Anonymous Cryptocurrency
In online shopping, consumers often rely on information such as sales, reviews or ratings to inform their decision making. Such preferences or user behaviors can be subjected to manipulation. For example, a merchant can artificially inflate product sales by paying a click farm. Specifically, the click farm will recruit a number of non-genuine buyers to purchase the products. After the purchases have been made, the buyers will either refund the product minus the commission or no product exchange actually takes place and these buyers are paid a commission for their role in the activity. Increasingly due to the popularity of cryptocurrency, such as bitcoin, such payment mechanisms are used in such activities. Hence, in this paper, we seek to detect click farm transactions using cryptocurrency. Specifically, we propose three models to capture click farm operations, and based on the models we design three algorithms to detect anonymous click farm transactions. Extensive analysis demonstrates that our model achieves a high accuracy rate in detecting anonymous click farm transactions, without incurring expensive computational costs.
Language : Python
Environment : Python3.7, Pytorch, Anaconda