Abstract
When reading or viewing long-form narrative media such as books, audio content, or television shows, or engaging in games with long gameplay durations, users may sometimes forget a particular detail about a character or entity in the content. While online resources as well as built-in features in the viewing application can help aid user recall, such resources are static and do not account for the user’s current progress, thus risking spoilers. This disclosure describes techniques that enable users to obtain entity information for narrative media in a spoiler-free manner. Per the techniques, a structured database of various entities in narrative media and facts about the entities is created, with individual facts associated with a progress point such as page number, a time stamp, etc. When the user initiates a query for an entity, for example by selecting the entity's name or asking a question, the facts about the entity from the database that are prior to the user’s current progress are retrieved. A generative artificial intelligence model is utilized to synthesize a response based on the retrieved facts.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Molina, Alejandra; Tsai, Vivian; and Jiang, Ellen, "Providing Spoiler-free Entity Information for Narrative Media Based on Viewing Progress", Technical Disclosure Commons, (September 29, 2025)
https://www.tdcommons.org/dpubs_series/8646