Abstract
Mobile device applications and operating systems offer a data-saving option such that online content that is believed to be unchanged since last load is not reloaded. However, this sometimes leads to a situation where online content that has actually changed is not reloaded or that stale content is reloaded. The misdetection of fresh content as stale or vice-versa occurs due to the heuristics that are used by the data-saving algorithms of the application/ OS, due to misconfigured servers, etc.
This disclosure presents machine-learning techniques that determine if a web page or a portion thereof is to be reloaded. The techniques use various contextual signals, as permitted by the user, to make the reload decision, e.g., the content to be reloaded; surrounding content; their metadata and positions on a web-page; user interaction and behavior with the website; previously loaded content; etc. The techniques enable data-saving techniques that are robust and tailored to both user and website.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Carbune, Victor and Coppey, Thierry, "Data Saving Using Contextual Signals", Technical Disclosure Commons, (October 15, 2018)
https://www.tdcommons.org/dpubs_series/1592