Google May Get More Deeply Personal

August 2, 2009

Google has a number of clever innovations in its personalization systems. The New York Times has discovered that can create a personalized television station. That’s just the tip of the personalization iceberg at Google. This iceberg, however, broke off quite a while ago and just now has collided with those lovable Titanics of mainstream media. Here are three patent patent applications that indicate the depth of the Google’s “for you, right now, no human editors needed” approach to personalization of network-centric information.

Many people worked on these. The USPTO, prompt as always, published three patent applications with the handiwork of Ashutosh Garg evident. Dr. Garg was, according to an interview in a Beckman Institute online publication:

At Google, I was the architect of the largest online personalization system.—Ashutosh Garg

Dr. Garg has left Google to pursue other interests. The purpose of this post is to capture the moving parts of Google’s potential in personalization of search and other services. The patent documents cited below provide an indication of the work that Microsoft and Yahoo will have to do to match Google’s “tailoring” capability. Killing Google will take more than media coverage, a 10 year deal, and assuming Yahoo’s punishing search cost burden.

Google has been working in this field for several years. You can find some historical color here.

20090192986, “Providing Content Using Stored Query Information”

This document was filed in January 2008. This system and method cranks through a user’s search history
and makes decisions about what to display.

The abstract said:

Among other disclosed subject matter, a computer-implemented method relating to providing content on a page includes receiving information for providing content for an access device in response to a navigation from a first page to a second page. The content is to be included in the second page. The method includes accessing query information stored on the access device, the query information based on a first query that was submitted from the access device to a search provider before navigating to the first page. The method includes providing the content for inclusion in the second page, the content selected using at least the received information and the accessed query information. A computer-implemented method can include recording search query information for a user; retrieving the search query information; and using the search query information to determine content for display to the user.

20090192888, “Targeted Ads Based On User Purchases”

The system and method is applied to a user’s purchase history. The idea is to suggest what other goodies the Google user will crave:

The present invention relates to systems and methods for providing advertisements on websites. In an embodiment, a method for providing an advertisement on a website includes obtaining purchase information submitted by a user making a purchase on the website, determining at least one advertisement for a product or service related to the purchase but of a different type than the purchase, and displaying the at least one advertisement on the website when the purchase is completed. In another embodiment, a system for providing an advertisement on a website includes a purchase server, an advertisement source, an analyzer, and an advertisement server.

This patent application was also filed in January 2008.

20090192705, “Adaptive and Personalized Navigation System”

This patent application was filed in March 2009, but seems to be conceptually related to the other two patent applications mentioned in this document. The abstract said:

Adaptive navigation techniques are disclosed that allow navigation systems to learn from a user’s personal driving history. As a user drives, models are developed and maintained to learn or otherwise capture the driver’s personal driving habits and preferences. Example models include road speed, hazard, favored route, and disfavored route models. Other attributes can be used as well, whether based on the user’s personal driving data or driving data aggregated from a number of users. The models can be learned under explicit conditions (e.g., time of day/week, driver ID) and/or under implicit conditions (e.g., weather, drivers urgency, as inferred from sensor data). Thus, models for a plurality of attributes can be learned, as well as one or more models for each attribute under a plurality of conditions. Attributes can be weighted according to user preference. The attribute weights and/or models can be used in selecting a best route for user.


These inventions embed adaptive personalization in the Google data centers and global computing fabric. The notion of personalization is that the system will figure out what a user wants and needs. The approach is reasonably clever. Based on my research into Google’s technology, the appearance of several related patent applications may indicate that there will be more personalization functionality available at any time.

Stephen Arnold, August 3, 2009


Comments are closed.

  • Archives

  • Recent Posts

  • Meta