CoreMedia Adaptive Personalization provides everything you need to implement rule-based personalization for your website out of the box. But what does this mean, exactly? Here's a simple example:

Assume you're the editor of a news site and there's a single main teaser region on your entry page. You know that placing a relevant teaser in this region is critical as it drives a high percentage of clicks, and more clicks mean more revenue. By inspecting the reports of your analytics system, you've noticed that in the morning, most visitors read World News, while during lunch break and in the evenings, interests are more diverse. In particular, you see some visitors focusing on Lifestyle, others on Economy, and still others on Sports. You decide to optimize your entry page by placing a Personalized Content in the main teaser region. This content item is configured to show the most important (as defined by the editorial team) article in World News each morning until 10am. At 10am, it switches to another Personalized Content that selects the most important teaser from Lifestyle, Economy, or Sports depending on the interests of the current user. If CoreMedia Adaptive Personalization is installed in your site, you can do all of this without the need to do any programming or to redeploy the system.

Figure 3.1. Example Page with Main Teaser

Have a quick look at the components of CoreMedia Adaptive Personalization that would be used to implement a system that supports this scenario. Detailed descriptions of the components can be found in the corresponding chapters within this manual.

All context data for a request, containing the current time of day, is stored within the CAE in a ContextCollection. If you tag your pages with keywords, as is typically the case if you use an ad server, you can use a ScoringContext in combination with the KeywordInterceptor to track the most often seen keywords for each user (if he read a lot of articles tagged with 'Sports', 'Sports' will have a large score in the context).

Personalized Content contains a list of rules of the form select content X if the contexts satisfy conditions Y. Given a ContextCollection, it renders the first content for which the conditions are satisfied. So in the scenario above, the main teaser would contain the rules select most-important-world-news if current time < 10am and select special-interest-article if current time >= 10am. The content most-important-world-news could use a search to determine the most current, highly rated editorial article from World News, while special-interest-article would be another Personalized Content selecting articles based on the users' keyword scores, for example select most-important-sports-news if score of 'Sports' > 0.8. These Selection Rules are defined from within CoreMedia Studio using a specialized editor component and are deployed to the CAEs via content item publication, so there's no need for any code changes.