Back in September, 2001, Google acquired the technological assets of Outride, which specialized in online information retrieval technologies. A white paper from the Outride group explains fairly well one of the approaches that they were taking in the field of personalized search (pdf).
We posit that at least two different computational techniques need to be combined to personalize search: contextualization and individualization. By contextualization, we mean the interrelated conditions that occur within an activity. Individualization means the totality of characteristics that distinguishes an individual. Contextualization includes factors like the nature of information available, the information currently being examined, the applications in use, when, and so on. Individualization encompasses elements like the user’s goals, prior and tacit knowledge, past information-seeking behaviors, among others. These elements are used to build a user model to personal relevancy computationally, as we will describe. It is this focus on the user and their context within the application of search that makes personalized search a compelling area to explore within the framework of contextual computing.
Google has developed some of its personalization technology, but the research and methods developed by Outride may have played a part in what they’ve done so far. I thought that this was an interesting set of comments in the paper from Outride:
It is worth mentioning upfront that since the following techniques alter the search experience, careful integration of these features into the user interface is required. In particular, the interface needs to provide a way to explain what the system is doing to personalize the experience as well as to undo the personalization.
I’ve had a couple of people ask me why they’ve seen some odd features in Google’s results, such as numbers listed after some results. The numbers they saw indicated the number of times they had visited that page before – and these folks hadn’t realized that they had personalized search turned on. That makes me wonder if Google needs to make it more obvious that personalized search is on, and that it is reranking and altering results based upon personalization.
Some of the technologies in the paper from Outride are discussed much more in detail in a new patent granted to Google last week.
Interface and system for providing persistent contextual relevance for commerce activities in a networked environment
Inventors: Donald R. Turnbull and Hinrich Schuetze
Assigned to Google
US Patent 7,089,237
Granted August 8, 2006
Filed January 26, 2001
Abstract
A search and recommendation system employs the preferences and profiles of individual users and groups within a community of users, as well as information derived from categorically organized content pointers, to augment electronic commerce related searches, re-rank search results, and provide recommendations for commerce-related objects based on an initial subject-matter query and an interaction history of a user. The search and recommendation system operates in the context of a content pointer manager, which stores individual users’ content pointers (some of which may be published or shared for group use) on a centralized content pointer database connected to a network. The shared content pointer manager is implemented as a distributed program, portions of which operate on users’ terminals and other portions of which operate on the centralized content pointer database. A user’s content pointers are organized following a local topical categorical hierarchy. The hierarchical organization is used to define a relevance context within which returned objects are evaluated and ordered.
The technologies described in the patent could be used both for personalizing organic search results and for providing eCommerce recommendations. As I was reading through it, I was asking myself, “How much of this might Google use,” and “How much have they already incorporated into their personalization efforts.” The document describes a vision of personalization that is more Outride’s than Google’s yet it’s interesting as a very detailed description of an approach to personalization.
Ed Sim, who was part of Outride, wrote a blog post about Google’s personalization, and the role of technology from Outride within it a couple of years ago. His assessment then seems to be that Google decided to follow personalization methods developed by another acquisition, Kaltix, which focused more on user profiles than user behavior. Several recent patent applications and papers from Google, such as Retroactive Answering of Search Queries, do show that user behavior isn’t something that Google is ignoring.