Compound ads present multiple items (e.g., houses, cars, jobs) together. A compound ad has two advantages:
it is larger, so it draws the eye
it requires fewer lines than separate ads would, because heading and contact information
are not repeated.
Compound ads were discussed at the task force meeting in Denver on August 7, 1998. Based on the discussion at the meeting and some subsequent thoughts, we would like to propose the following approach:
Require that an ad body consist of one or more segments. Compound ads would have more than one segment.
Allow each segment to be coded as type="common" or type="item" (default). An example of a “common” segment could be the Car dealer and contact information. An example of a “item” segment could be a particular car in an ad with one or more cars.
Include both the text element and (when available) the coding1 element in each segment.
Allow segments to nest. This allows common text to bond to some, but not all items.
For print: concatenate the segments in their document order.
For electronic media:
use each segment’s coding for search,
reproduce the entire ad like printORsplit off each item segment as a separate ad,
repeating all “common” containing segments for each.
1 Coding is a rich, structured representation of the ad content. Coding also supports ad searching. Finally, where the classification handling is insufficient to identify the ad class, coding assists downline users in classifying the ad. The values inside the <coding> </coding> element are standardized for data storage and comparison, not formatted for presentation. For example, the telephone
number inside the <phone> </phone> element is represented as a string of digits without punctuation. If the same information is also included in the text element, it would be formatted for presentation there.