Abstract
Browser-native content adaptation modifies web-page content represented in a Document Object Model (DOM) based on contextual information indicating a user’s familiarity with the subject matter of the content. A user expertise profile is generated from historical interaction data including browser history, searches, pages read, semantic exposure, frequency of encounters, dwell time, and viewport activity. Before or during rendering of a target page, terms, acronyms, sentences, images, and other content are evaluated against the profile and/or page context. Content determined to be unfamiliar to the user is adaptively modified by, for example, expanding acronyms, adding contextual explanations, simplifying language, or injecting another comprehension aid. Content and the user’s contextual information disambiguate expressions having different meanings in different domains. Adaptation can be performed locally using an on-device language model and can be inserted through a layout-preserving rendering layer. User interaction with adapted content can be used to update the profile for later browsing sessions.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Prakash, Abhay, "Dynamic Browser-Native Content Simplification and Contextual Injection Based on Historical Interaction and Localized Expertise Modeling", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/11811