Google is now attempting to predict what you’re searching for before you’re finished searching. On December 22nd, the world leading search engine filed a patent application for its technology used by Google Suggest, which analyzes keystrokes and brings back a list of suggested terms in real time before a query is fully entered into the search bar.

According to Google, the idea behind Google Suggest is to make search queries faster, more relevant and personalized for the user. The technology works by sending a partial query to the engine as you begin typing, which then returns a dynamic list of terms gathered from a dictionary. It displays these terms in real time under the search bar in an endeavor to guess what you’re looking for. For example, while entering ‘prog’ into the search bar, the first three terms the algorithm returns include: progressive, progressive insurance and progeria. These top three terms then change to program files, programming, and programz when the ‘r’ is entered after the ‘g.’

The patent claims the technology works in a number of possible ways. Suggestions may stem from prior searches entered by a community of users, the current user’s profile, or cached search terms. Bill Slawski at Cre8asite Forum believes these terms originate from most popular searches, recent searches, hot topics, or cached terms. His blog on the subject can be found here.

Although Google’s technology may be unique, it isn’t a new idea. There are various search engines that incorporate similar technologies. For instance, Become, an ecommerce centric engine, also makes dynamic suggestions as you type into the search bar via its ‘Dynamic Search Term’ technology.

The inventors on the Google application include: D. Kamvar D; H. Haveliwala; and M. Jeh. Google Suggest is currently still in Beta.