US Patent 11263031 Assisting users with personalized and contextual communication content

In one embodiment, a method includes receiving a first user input by a first user at a client system associated with the first user, wherein the first user input is associated with a first language register, selecting a first language-register model from a plurality of language-register models based on the first language register by the client system, wherein the plurality of language-register models are each personalized with respect to the first user, and presenting a personalized first communication content generated based on the selected first language-register model at the client system.