In a world where searches are pushing for greater personalization, Bing has stepped up their game via Matchbox. Matchbox, a technology that uses multiple forms of input to bring “likable” content to a user, brings personalization to a whole new level.
How Non-Matchbox Personalization Works
There’s little doubt that personalization is of great importance in the modern internet. As the web continues to evolve, it’s becoming more and more clear that it’s a “social scene.” Even on the web, no person is an island, and with the 21st century expansion of social networks as a primary medium of interaction, more groups and turning toward their friends, followers, and connections to find valuable sites and sources. Social interaction is seen both on social sites and as a way to help personalize results in search, but that’s not the only way that results pages will be personalized.
Users who login and have information stored will also be subject to “collaborative filtering.” This filtering process lumps people together based on their interests. When a user or set of users from this group interacts with a site, the filter can assume that their interest has a greater likelihood to be shared by others in the same group. Users may be familiar with this concept through services such as Pandora Radio or Netflix, where the ratings of “similar” users provide the basic mechanism for recommendations.
Bing’s New Step Forward
The Bing Blog highlights the remaining question: “How do you approach personalization in a reality where there will not be enough activity to enable collaborative filtering systems to kick in?” This is especially applicable in the “real-time” search world, which includes items like breaking news.
Enter Matchbox, the solution Bing presents for this problem. Bing gave full details in their presentation at SXSW, but the brief version is as follows: The technology is designed to understand the subject-matter of content on the web, allowing Matchbox to see patterns in the behavior of users. If you like content that fits into a specific category and style, you’re likely to also enjoy other content from a similar category and style.
As the first major “Matchbox project,” Bing is releasing Project Emporia. Emporia, a web and mobile phone app, uses multiple filters for personalization, including Matchbox, social recommendations (primarily via Twitter), and collaborative filtering. More exciting, however, is the potential impact of Matchbox on actual Bing searches; users can expect more information on this front in the near future.
[via the Bing Blog]