Speculative actions based on user dwell time over selectable content

摘要

Methods, systems, and techniques for reducing web page load time are provided. An exemplary method for reducing web page load time includes detecting, in a browser, a cursor in a region of a web page. The region surrounds a selectable item associated with a URL. The exemplary method includes determining whether the cursor is present in the region for a threshold period of time. The exemplary method further includes performing a speculative action when it is determined that the cursor has been present in the region for the threshold period of time, without modifying underlying code of a web page associated with the URL.

主权项

1. A method for reducing web page load time, comprising:
detecting, in a browser, a cursor in a region of a web page, the region surrounding a selectable item associated with a URL; receiving speculative information from a site based on a popularity of the URL; determining whether the cursor is present in the region for a first threshold period of time, wherein the first threshold period of time is dynamically determined based on the received speculative information; performing a first speculative action when it is determined that the cursor has been present in the region for the first threshold period of time, without modifying underlying code of a web page associated with the URL, wherein the first speculative action comprises one of: performing a first DNS preresolution of the URL, initiating a first preconnection to the URL, and prefetching a first content of the URL; determining whether the cursor is present in the region for a second threshold period of time greater than the first threshold period of time; and performing a second speculative action, different from the first speculative action, when it is determined that the cursor has been present in the region for the second threshold period of time, wherein the second speculative action comprises one of: performing a second DNS preresolution of the URL, initiating a second preconnection to the URL, and prefetching a second content of the URL.