An intelligent approach to generate personalized hints in a serious game

• Main Author: Lin, Zichun
• Other Authors: Miao Chun Yan
• Format: Thesis-Master by Research
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Engineering::Computer science and engineering
• Online Access: https://hdl.handle.net/10356/159794
• Institution: Nanyang Technological University

With the development of Internet Technology and Artificially Intelligence, online learning is gaining popularity. Students may encounter resources like video lectures, reading content, quizzes, discussion forums, and games in online learning. As one of the most efficient and engaging methods of online learning, games naturally get more attention than other methods, especially among children. This thesis proposes an intelligent approach for generating personalized hints in a serious educational game. There are virtual experiments, reading contents, and multiple-choice questions in the game. A Reinforcement Learning model could be built to generate next-step hints to assist students in playing the game. However, the generated raw hints will not be displayed to the player directly, which need to be processed again. Each player’s Felder-Silverman learning style is calculated based on a pre-designed questionnaire. A hint library was created in order to personalize hints based on the player’s learning style, which may inspire their curiosity. The final hint is chosen from the library. To make learning more enjoyable and to provide learning companionship, the TmallGenie Smart Speaker is used to communicate with players. In addition to displaying text hints in the game, hints may be played by the speaker using its pleasant voice. It can also read aloud the introduction contents and respond to simple questions from the player. According to results, the majority of students agreed that the game with personalized hints improves their knowledge comprehension and makes them feel better. Combining the game with TmallGenie also provides them a lot of excitement and novelty. However, the result has some limitations. Due to insufficient students participated in the experiment, the experiment’s outcome may not be particularly compelling. And because the learning content is not complicated, the quizzes didn’t distinguish the learning outcome with good precision. In conclusion, by playing the educational game, students can learn the knowledge in a more personalized way, achieving both personalized learning and education through entertainment.