| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 5.34 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
Sistemas de Recomendação ajudam os seus utilizadores ao servir recomendações com base nos
seus gostos. Este tipo de sistema abrange vários campos como entretenimento, produtos, e até
pessoas. O uso de personalidade tem ganho popularidade como uma forma de produzir
recomendações de alta qualidade e precisão. No entanto, estes sistemas sofrem da necessidade
de preencher questionários longos e intrusivos ou de minerar grandes quantidades de dados
para captar a personalidade dos utilizadores.
Este trabalho propôs o desenvolvimento de um jogo de curta duração para dispositivos móveis,
para tentar captar o traço de personalidade ansiedade, de forma não intrusiva, para integração
num sistema de recomendações. Após recolher informação sobre trabalhos relacionados e
sobre as áreas de estudo relevantes, optou-se por desenvolver um jogo do género survival
horror em Unity, para dispositivos Android.
De forma a validar se o jogo estava a medir o traço de personalidade pretendido, foram
conduzidas experiências com 134 participantes voluntários, dos quais 93 foram considerados
viáveis. No final não foi possível estabelecer correlações significativas entre os valores medidos
pelo jogo e os valores medidos pelo questionário para o traço ansiedade. Presumidamente,
devido às condições de conduta do teste não terem sido imersivos o suficiente para o tipo de
experiência que se esperava que o jogador tivesse ao jogar o jogo. Foram encontradas, mesmo
assim, algumas correlações significativas com outros traços de personalidade como cooperação
e altruísmo.
Recommender Systems assist their users by serving recommendations based on their tastes and preferences. This type of system spans several areas such as entertainment, products, and even people. The use of personality as a way to produce more accurate, higher quality recommendations has seen increased use over time. However, these systems suffer from the need to fill long and intrusive queries or mining large datasets as the only ways to acquire user personality. This paper proposes the development of a short mobile game to try to capture the anxiety personality trait, in a non-intrusive manner, for integration in a recommender system. After gathering the necessary information on related works and relevant domains of study, it was decided to develop a survival horror game in Unity, for Android devices. In order do validate if the game was measuring the intended personality trait, practical experiments were conducted featuring 134 voluntary participants, 93 of which were considered viable. In the end it was not possible to establish correlation between the values measured by the game and the personality test for the anxiety trait. This was presumably due to the playtesting conditions not being immersive enough for the type of experience the player was intended to undergo. Even still, significant correlation was established with other personality traits, such as cooperation and altruism.
Recommender Systems assist their users by serving recommendations based on their tastes and preferences. This type of system spans several areas such as entertainment, products, and even people. The use of personality as a way to produce more accurate, higher quality recommendations has seen increased use over time. However, these systems suffer from the need to fill long and intrusive queries or mining large datasets as the only ways to acquire user personality. This paper proposes the development of a short mobile game to try to capture the anxiety personality trait, in a non-intrusive manner, for integration in a recommender system. After gathering the necessary information on related works and relevant domains of study, it was decided to develop a survival horror game in Unity, for Android devices. In order do validate if the game was measuring the intended personality trait, practical experiments were conducted featuring 134 voluntary participants, 93 of which were considered viable. In the end it was not possible to establish correlation between the values measured by the game and the personality test for the anxiety trait. This was presumably due to the playtesting conditions not being immersive enough for the type of experience the player was intended to undergo. Even still, significant correlation was established with other personality traits, such as cooperation and altruism.
Description
Financiado
Keywords
Serious game Personality trait Anxiety Recommender system Jogo sério Ansiedade Personalidade Sistema de recomendação