Our research introduces STORY2GAME, a groundbreaking approach that leverages Large Language Models to generate text-based interactive fiction games. This innovative method begins by crafting a narrative and constructing the game world, followed by coding actions in a game engine to enable interactive storytelling. Unlike traditional approaches that rely on predetermined actions, our method allows for more open-ended story generation while ensuring experiences remain rooted in the game state. We delve into the success rates of action code generation and highlight the percentage of successful actions and the number of stories where all actions were executed seamlessly. Our findings show that most sentences compile successfully, with occasional challenges arising from object misidentification by the LLM. This issue typically occurs when objects are described with varying adjectives, leading to ambiguity in their identification within the game state. Furthermore, we explore dynamic action generation and evaluate the completeness and coherence of newly generated actions. By prompting the LLM to provide verbs for items and characters in a story, we ensure that these verbs are unique and enhance player interactions within the game. Our analysis reveals that new items are commonly used as preconditions for actions due to their integral role in gameplay mechanics. Additionally, we examine how different types of preconditions—such as creating new objects or attributes—impact action generation for characters and items in a story. We observe varying trends in precondition selection based on the nature of interactions with characters versus objects, shedding light on how novel actions can be executed within the game environment. Overall, our research showcases the potential of using Large Language Models for generating interactive fiction games with dynamic action capabilities. By refining our approach to address challenges like object misidentification and optimizing precondition selection for diverse gameplay scenarios, we aim to enhance player engagement and create immersive story-driven gaming experiences.
- Introduction of STORY2GAME: a new approach using Large Language Models for text-based interactive fiction games
- Process involves crafting narrative, constructing game world, and coding actions in a game engine for interactive storytelling
- Emphasis on open-ended story generation while maintaining connection to game state
- Success rates of action code generation analyzed, focusing on successful actions and seamless execution in stories
- Challenges with object misidentification by LLM due to varying adjectives used for objects in descriptions
- Exploration of dynamic action generation by prompting LLM to provide unique verbs for items and characters
- New items commonly used as preconditions for actions to enhance player interactions within the game
- Examination of how different types of preconditions impact action generation for characters and items, highlighting trends based on interaction nature
- Potential of Large Language Models in creating immersive story-driven gaming experiences with dynamic action capabilities
- STORY2GAME is a new way to make interactive stories using big language models.
- It involves making up a story, creating a game world, and putting actions into the game for storytelling.
- The focus is on making stories that can change based on what happens in the game.
- They looked at how well the game can understand and do the actions you tell it to do.
- They found it can be tricky when the game gets confused about objects because of different words used.
Definitions
- Interactive fiction games: Games where players make choices that affect the story's outcome.
- Narrative: A story or account of events.
- Game engine: Software used to create video games.
- Action code generation: Making instructions for what happens in a game based on player input.
- Adjectives: Words that describe nouns (objects).
- Verbs: Words that show action or state of being.
- Preconditions: Conditions that must be met before an action can take place.
- Immersive: Providing deep mental involvement in something.
Introduction:
Interactive fiction games have been around for decades, captivating players with their immersive storytelling and open-ended gameplay. However, the traditional approach to creating these games often relies on predetermined actions, limiting the potential for dynamic storytelling and player engagement. Our research paper introduces a groundbreaking method called STORY2GAME that leverages Large Language Models (LLMs) to generate text-based interactive fiction games. This innovative approach allows for more open-ended story generation while ensuring experiences remain rooted in the game state.
Crafting the Narrative and Constructing the Game World:
The first step in our method is crafting a narrative and constructing the game world. This involves creating a detailed storyline with well-developed characters, settings, and plot points. The LLM then uses this narrative as a base to generate text-based descriptions of scenes, objects, and characters within the game world.
Coding Actions in a Game Engine:
Once the narrative and game world are established, we move on to coding actions in a game engine to enable interactive storytelling. Unlike traditional approaches that rely on predetermined actions, our method allows for more flexibility by prompting the LLM to provide verbs for items and characters in a story. These unique verbs enhance player interactions within the game and create dynamic action capabilities.
Success Rates of Action Code Generation:
Our research also delves into success rates of action code generation using LLMs. We highlight both the percentage of successful actions as well as the number of stories where all actions were executed seamlessly. Our findings show that most sentences compile successfully; however, there are occasional challenges arising from object misidentification by the LLM.
Challenges with Object Misidentification:
Object misidentification occurs when objects are described with varying adjectives leading to ambiguity in their identification within the game state. To address this issue, we continue refining our approach by training LLMs on larger datasets containing diverse descriptions of objects.
Dynamic Action Generation:
One of our key focuses is on dynamic action generation, where the LLM generates new actions based on prompts from the game engine. We evaluate the completeness and coherence of these newly generated actions to ensure they align with the overall narrative and enhance player engagement.
Precondition Selection for Diverse Gameplay Scenarios:
We also examine how different types of preconditions impact action generation for characters and items in a story. Our analysis reveals varying trends in precondition selection based on the nature of interactions with characters versus objects. This sheds light on how novel actions can be executed within the game environment, providing insights into creating diverse gameplay scenarios.
Conclusion:
Our research showcases the potential of using Large Language Models for generating interactive fiction games with dynamic action capabilities. By refining our approach to address challenges like object misidentification and optimizing precondition selection for diverse gameplay scenarios, we aim to enhance player engagement and create immersive story-driven gaming experiences. With further advancements in LLM technology, we believe that STORY2GAME has immense potential to revolutionize the world of interactive fiction games and provide players with endless possibilities for storytelling and gameplay.
