Philippine E-Journals

THEY KNOW: A PATHFINDING RATIONAL AGENT VIA KNOWLEDGE ACQUISITION MODELING FROM PLAYER BEHAVIOR AND ENHANCED A* ALGORITHM

Joshua Rei Aviles | Christopher John Hispano | Jaen Dave Lucas | Edwin M Torralba

 

Abstract:

Modern games have used Artificial Intelligence as a core element of game design and mechanics. Artificial Intelligence has thrived in many fields; however, the enhancement of pathfinding algorithms of AIs in games have remained stagnant. Moreover, numerous issues with regards to the exploitable plot of games have been raised by several gamers and game developers. The purpose of this research is to develop They Know - a survival themed thriller game with 4J MT-Adaptive A* - an adaptive and enhanced pathfinding algorithm created using A* as basis, to reduce the predictability of in-game agents by using the developed algorithm and by modifying agents with knowledge acquisition, to distinguish the personality and emotion of modern players based on the reaction of players to in-game situations, and to acquire data for determining the motivation of contemporary gamers for playing. The efficiency of the developed algorithm was compared with Lazy MT – Adaptive A* developed by Koenig (2007). Questionnaires developed by HEXACO, Power (2017), and Demetrovics (2011) were used as standards for acquiring data from the respondents to determine the respondents’ personality, degree of predictability of the game, and the motivation of players in playing. Data were gathered from grades 11 and 12 students of the University of Santo Tomas – Senior High School.