Mental health challenges, including depression, anxiety, and eating disorders, are increasingly affecting young people, highlighting a critical need for effective monitoring and early intervention strategies. Traditional assessment methods often lack the ecological validity to capture the complexities of real-world motivational and decision-making processes. This work proposes a novel computational framework centered on a gamified task to measure emotion-sensitive behavioral metrics, particularly focusing on extrinsic motivation. We present the design specifications of a mobile video game, developed for the Android operating system, that simulates a fruit-collecting foraging task. Drawing on foraging theory, Charnov’s Marginal Value Theorem, the free energy principle, and active inference, the game is structured to elicit goal-directed behavior under uncertainty and to model player actions computationally. The design prioritizes simplicity and user-friendliness, employing a simple tap-based interaction and a positive reward system, ensuring accessibility regardless of user skill level. During game-play, the system records player interactions as sequences of progress states, capturing variables such as path taken, energy expenditure, time management, and fruits collected. These data enable the extraction of individual-level parameters serving as functional markers of motivational strategy. By analyzing game-play patterns, distinct behavioral profiles indicative of various emotional states—such as apathetic/depressed, greedy strategist, anxious but semi-strategic, dependent, and mixed/unstable—can be identified. This model-based approach offers a dynamic alternative to conventional self-report methods for characterizing behavioral phenotypes. The proposed framework provides a feasible and attractive tool for psychologists and psychiatrists to monitor their younger patients, aiding in the identification of early behavioral markers of vulnerability and supporting longitudinal clinical monitoring. Future work includes implementing the game as a mobile application and validating it with real participants.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

A Computational Framework for Emotion-Sensitive Behavioral Metrics Using a Gamified Task

  • Diego Riofrío-Luzcando,
  • Miguel De Andrés,
  • Victoria López,
  • Matilde Santos,
  • Diego Urgelés

摘要

Mental health challenges, including depression, anxiety, and eating disorders, are increasingly affecting young people, highlighting a critical need for effective monitoring and early intervention strategies. Traditional assessment methods often lack the ecological validity to capture the complexities of real-world motivational and decision-making processes. This work proposes a novel computational framework centered on a gamified task to measure emotion-sensitive behavioral metrics, particularly focusing on extrinsic motivation. We present the design specifications of a mobile video game, developed for the Android operating system, that simulates a fruit-collecting foraging task. Drawing on foraging theory, Charnov’s Marginal Value Theorem, the free energy principle, and active inference, the game is structured to elicit goal-directed behavior under uncertainty and to model player actions computationally. The design prioritizes simplicity and user-friendliness, employing a simple tap-based interaction and a positive reward system, ensuring accessibility regardless of user skill level. During game-play, the system records player interactions as sequences of progress states, capturing variables such as path taken, energy expenditure, time management, and fruits collected. These data enable the extraction of individual-level parameters serving as functional markers of motivational strategy. By analyzing game-play patterns, distinct behavioral profiles indicative of various emotional states—such as apathetic/depressed, greedy strategist, anxious but semi-strategic, dependent, and mixed/unstable—can be identified. This model-based approach offers a dynamic alternative to conventional self-report methods for characterizing behavioral phenotypes. The proposed framework provides a feasible and attractive tool for psychologists and psychiatrists to monitor their younger patients, aiding in the identification of early behavioral markers of vulnerability and supporting longitudinal clinical monitoring. Future work includes implementing the game as a mobile application and validating it with real participants.