View in EDS

Development and Implementation of the Algorithm Designed to Struggle with Videogame Addiction

Saved in:
Bibliographic Details
Title: Development and Implementation of the Algorithm Designed to Struggle with Videogame Addiction
Authors: Dmitry Sychugov, Uliana Ilinykh
Source: Современные информационные технологии и IT-образование, Vol 18, Iss 1, Pp 202-211 (2022)
Publisher Information: The Fund for Promotion of Internet media, IT education, human development «League Internet Media»
Publication Year: 2022
Collection: Directory of Open Access Journals: DOAJ Articles
Subject Terms: computer games, video games, game addiction, struggling with the game addiction, key pressing dynamics, keyboard user input, keylogger program, software development, java development, neural network, machine learning, random forest classifier, Electronic computers. Computer science, QA75.5-76.95
Description: Game addiction appears to be a widespread disease among young people. Nowadays methods, such as parent control programs, have their own disadvantages. In this work a new method for controlling game activity is researched. It depends on analyzing user’s keyboard input. Research includes experiment in the order to collect information about key pressing dynamics during gaming. It turns out that gamers have their own unique pattern in how they press the keys. The most widespread type of the games are the games using four main buttons to control characters. Such games clearly differ from other types of computer activity such as typing or writing code. The main goal of this work is to develop and implement the algorithm which uses this pattern to detect games and stop computer process which runs the game. Machine learning methods appeared to be good for these purposes. The experiment allowed us to create a big database with input scans. Records are used to train a random forest classifier. Java programming language is used for implementing algorithm as a program product. JNativeHook open library is used for scanning process. At the start, the program creats a random forest classifier and trains it with static data. After this the main cycle of the program starts: every 20 seconds a new scan is performed and analyzed. If the program determines that there is currently game activity, then it will try to detect the game computer process and to stop it. During testing program showed great results for game with described above pattern: it detected the game and stopped its process for ten times in a row without any false effects.
Document Type: article in journal/newspaper
Language: Russian
Relation: http://sitito.cs.msu.ru/index.php/SITITO/article/view/833; https://doaj.org/toc/2411-1473; https://doaj.org/article/43bb3eee3bf3469e85ce8d42d563aaf4
DOI: 10.25559/SITITO.18.202201.202-211
Availability: https://doi.org/10.25559/SITITO.18.202201.202-211
https://doaj.org/article/43bb3eee3bf3469e85ce8d42d563aaf4
Accession Number: edsbas.C2EE7F96
Database: BASE
Description
Abstract:Game addiction appears to be a widespread disease among young people. Nowadays methods, such as parent control programs, have their own disadvantages. In this work a new method for controlling game activity is researched. It depends on analyzing user’s keyboard input. Research includes experiment in the order to collect information about key pressing dynamics during gaming. It turns out that gamers have their own unique pattern in how they press the keys. The most widespread type of the games are the games using four main buttons to control characters. Such games clearly differ from other types of computer activity such as typing or writing code. The main goal of this work is to develop and implement the algorithm which uses this pattern to detect games and stop computer process which runs the game. Machine learning methods appeared to be good for these purposes. The experiment allowed us to create a big database with input scans. Records are used to train a random forest classifier. Java programming language is used for implementing algorithm as a program product. JNativeHook open library is used for scanning process. At the start, the program creats a random forest classifier and trains it with static data. After this the main cycle of the program starts: every 20 seconds a new scan is performed and analyzed. If the program determines that there is currently game activity, then it will try to detect the game computer process and to stop it. During testing program showed great results for game with described above pattern: it detected the game and stopped its process for ten times in a row without any false effects.
DOI:10.25559/SITITO.18.202201.202-211