Caution: This documentation is for eZ Publish legacy, from version 3.x to 6.x.
For 5.x documentation covering Platform see eZ Documentation Center, for difference between legacy and Platform see 5.x Architecture overview.

Content Export Mapping

Content export mapping

We stated earlier that the recommendation service is a cloud-based service. It uses pre-calculated models that are based not only on the user´s activities but also on information about the content of your eZ Publish installation.

Each time you publish some content and the content is marked with ´Export content (for recommendation)´ in the class definition, the extension sends the defined meta-data to the recommender engine in the cloud. With the content export mapping you define which information is sent to the recommender engine/used for future model calculation.

There are 10 content attributes that are extremely valuable for the calculation of recommendations and filtering of the results. You find the keys to these attributes on the left-hand side of the attribute mapping in your class definition. Two of these attributes are mandatory, but exporting more content information may significantly improve the quality of the results.

Example:

When your individual article class has a number of date/time entries like publish date, unpublish date, created on, validated on it is important to tell the recommender engine the time window for recommendation of each article (probably between publish date and unpublish date). In this case you map publish date with Valid from and unpublish date with Valid to).

On the right-hand side you can choose, from a selection box, which of your class attributes you want to map to these recommendation keys.

You may have additional content information that is not interpreted by default, like the color or size of a product, the category of an article or the resolution of an image. Such information can be useful to create filters on recommendation results prior to presentation.

In the Advanced version of the eZ Recommendation service you can make use of the additional information for such filter purposes. The Recommendation datatype therefore allows defining Additional attributes for content export.

The user interface shows all class attributes on the list. Please do not check twice the attributes you have already mapped to other attributes above. This creates more traffic from redundant content export information only.

Andrea Melo (30/05/2012 12:40 pm)

Andrea Melo (30/05/2012 12:40 pm)


Comments

There are no comments.