Pengembangan Sistem Pembangkit Cerita Interaktif pada Aplikasi Permainan Horor Psikologis Menggunakan Large Language Model (LLM)

Abstract

Content development for games is one of the most difficult challenges in building games, especially horror games. This is due to the need to create a truly terrifying experience, which requires a perfect blend of narrative, atmosphere and gameplay. Narration in video games, especially horror games, is often static and also limited to scenarios that have been predetermined by the developer. In recent years, the use of Large Language Model (LLM) such as ChatGPT has become essential to everyday life. Large Language Model (LLM) is a type of artificial intelligence model that is trained on large amounts of text data to understand, manipulate, and generate natural language that makes it capable of generating human-sounding text, writing stories, and answering questions. What if the static story or narrative is changed into a dynamic one generated by the Large Language Model that can increase the replayability of the horror game. This research develops a Story Generation System in psychological horror games using the Large Language Model (LLM). The role of Large Language Model (LLM) here is to generate stories when players interact with objects in psychological horror games that are also built. Objects that can be interacted with by the player will send spatial data as context injection for the Large Language Model (LLM) which will help the model to understand where each object is located in the game. This research uses LLaMA 3.2-1B-Instruct as a Large Language Model that is integrated and will be run locally in the game. Testing in this research is done by giving questionnaires to 10 users to assess whether the Large Language Model can generate a story in psychological horror games and can also help players in achieving their objectives.