I had seen this db structure in other project, but didn't know that it was called EAV. For those who don't read php|architect, EAV describes a method of saving a large set of attributes, only some of which apply to an individual entity. Normally you would create a table, with a row for each entity and save each attribute in a column. With EAV you save each attribute as a row.

He suggests two ways to do the modeling - the more "common" way and an alternative way that splits up the data types to make querying simpler (into scalars and arrays). A sample database structure and example query for it are included.

The mental image that came to me when reading about this approach to data modeling was taking the traditional relational table and turning it on its head. [...] EAV actually has to circumvent, work around, or reimplement features that most mainstream database servers today provide "for free" to the traditional relational counterparts of EAV in order to get equal functionality.

He talks about features the structure has (including row-level referential integrity and schema metadata) and points out some of the other document-focused database systems like Lucene and CouchDB.

Check out the story in the June 2008 issue of php|architect for that cover story from Carl Welch.