Analiza vpliva težavnosti računalniške igre na izmerjene vrednosti fizioloških signalov

Knez, Timotej (author), Pejović, Veljko (mentor), Gjoreski, Martin (co-mentor)

Cilj diplomske naloge je ugotoviti, ali je mogoče med igranjem igre sproti zaznavati miselni napor igralca, kar bi nam omogočilo dinamično prilagajanje težavnosti igre glede na potrebe posameznega igralca. Za določanje miselnega napora smo poskusili uporabiti fiziološke signale, ki jih lahko merijo pametne naprave. Ta ideja je danes še posebej aktualna, saj so naprave, ki omogočajo merjenje tovrstnih signalov, vse bolj pogoste. Da bi odkrili, ali je takšno zaznavanje napora mogoče, smo zasnovali eksperiment, s katerim smo preverili, ali obstaja povezava med izmerjenimi biološkimi signali ter težavnostjo igrane igre. Za potrebe izvedbe eksperimenta smo izdelali Android aplikacijo, ki uporabniku omogoča igranje igre Kača, hkrati pa s pomočjo pametne zapestnice meri fiziološke podatke. Te podatke smo kasneje obdelali s pomočjo tehnik strojnega učenja in izdelali več modelov za napovedovanje težavnosti igrane igre na podlagi izmerjenih signalov. Rezultate eksperimenta smo preverili z nekaj dodatnimi poskusi, s katerimi smo potrdili, da obstaja povezava med izmerjenimi fiziološkimi signali in težavnostjo igre.

• Language: Slovenian
• Keywords: {računalniške igre, dinamična težavnost, biološki signali, mobilno zaznavanje, strojno učenje, kognitivna obremenjenost
• Work type: Bachelor thesis/paper
• Organization: FRI - Faculty of Computer and Information Science
• Year: 2019
• PID: 20.500.12556/RUL-110571
• COBISS.SI-ID: 1538399939
• Publication date in RUL: 17.09.2019

Secondary language

• Language: English
• Title: Analizing effect of computer game difficulty on biological signals

Abstract

The goal of this thesis is to determine whether it is possible to track a gamer's mental demand in real time. If possible, this would enable us to dynamically adjust the game's difficulty according to each player's unique needs. Biometric signals, measurable by smart devices, were used in order to determine mental demand. This is nowadays a popular approach as the devices able to measure such signals are getting more common. To determine whether measuring mental demand in such a way was possible an experiment was designed to verify the existence of a connection between the measured biological signals and the difficulty of the game. An Android application was designed for the needs of the experiment which allows the users to play the game Snake whilst measuring biometric data via a smart wristband. This data was later processed using machine learning algorithms to create various models of determining a game's difficulty based on the measured signals. The results of the experiment were verified with some additional experiments to confirm a link between the measured biometric signals and the difficulty of a game.

• Keywords: computer game, dynamic difficulty, biological signals, mobile sensing, machine learning, cognitive load