Abstract
Content personalization systems typically determine which content to show a reader. Once content is selected, however, it is generally presented in the same fixed form to everyone. This disclosure describes a system that instead personalizes where a narrative begins, based on an estimate of what an individual reader already knows about the subject, called a cognitive identity. The cognitive identity is built from the reader's prior exposure to content related to the subject, an optional self-declared familiarity level, and engagement signals such as re-reading or early exit from related content. Given one body of source facts, the system generates several candidate starting points, each assuming a different level of prior knowledge, and selects the candidate that best matches the reader's cognitive identity. A narrative is then generated from the selected starting point, using only the source facts and an approved set of background facts, so every version produced from the same source stays factually consistent. Confirmed feedback on how well the selected starting point matched the reader's actual knowledge is used to refine later selections.
Creative Commons License
This work is licensed under a Creative Commons Attribution-Share Alike 4.0 License.
Recommended Citation
INC, HP, "Cognitive Identity-Based Narrative Entry Point Selection for Personalized Content Generation", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/12242