Application and Feasibility Analysis of Proactive Anti-cheating Algorithm Based on ML

Ruixi Liu

Abstract


Nowadays, most Internet application contents based on modern Internet service providers are realized through human-computer interaction. Therefore, the application scenarios of HCI in life are very wide, and the visual operation path provides great system availability and user-friendliness. And its security is an important fulcrum for service providers (SP) to obtain benefits, or fund transactions between terminals. According to the trigger mechanism of cheating, it is mainly divided into exposure cheating, click cheating, and conversion cheating. Cheating will greatly pollute the network environment and it even violates laws and regulations. Therefore, ISPs are deploying and continuously improving anti-cheat algorithms and mechanisms during the service development stage. The location of these operating mechanisms is generally the network server used by the SP to provide services, and some algorithms and monitoring behaviors significantly increase the load of the server, thereby increasing service costs.

Based on machine learning (ML) technology, this paper introduces a feasibility analysis of anti-cheat algorithm for human operation based on object terminal, which is suitable for behavior recognition level and data level. The algorithm learns user behavior through their active participation. It finds the path of the request by combining with packet matching of abnormal behavior characteristic of cheating behavior updated regularly by the service provider. This algorithm also detects cheating behavior through local controller. When the two peers have the same or a high probability of prejudgment after calculation, the server is triggered to actively identify user behavior, using the trust profile to analyze and log the cheating path. The compatibility and friendliness of terminal equipment characteristics and hardware level with AI algorithms in the current environment is helpful to reduce the load caused by the server mounting of the ISP and costs.


Keywords


Human-Computer Interaction (HCI); Machine Learning (ML); Artificial Intelligence (AI); Anti-cheat; Algorithm; Localization

Full Text:

PDF

References


Bai W. Research on cheat warning model based on user behavior analysis [MSc thesis]. Chengdu: University of Electronic Science and Technology of China; 2018.

Ma R. Design and implementation of anti cheating system based on analysis of content and user behavior about social platform [MSc thesis]. Harbin: Harbin Institute of Technology; 2015.

Yang Z. Design and implementation of anti-cheat analysis model in O2O scenario [MSc thesis] (in Chinese). Harbin: Harbin Institute of Technology; 2016.




DOI: http://dx.doi.org/10.18686/esta.v7i2.141

Refbacks

  • There are currently no refbacks.