Page-level Content

Question

Is it possible to implement rules or configuration settings that would prevent the indexing of certain page level content items?
The condition I'm trying to remediate is a web source with a menuing system which results in "false" positives for certain SES query phrases that appear in the menuing text, but not in the main content area of the page.
The menu is identified with a unique div id, so I'm wondering if I can exclude this portion of the page from indexing by implementing a rule or setting that targets a div id.
Alternatively, if there's another workaround for this sort of condition I'd be interested to hear about it.
I have read through both the Administrator's Guide and this forum, but can't find anything on the topic. Any help (even a definitive "no") will be useful.

Answer 1

This can be done from a "Document Service".
A document service has access to the entire document content, and can modify it as required. Whatever it writes back will be what gets indexed.
So your document service could look for particular patterns within the content and remove them (I'd suggest using the java regexp package).
Ideally the patterns to remove would be obtained from a config file so you don't need to keep recompiling the document service to deal with new patterns.
There is a sample document service that you can modify on our [OTN Page|http://www.oracle.com/technology/products/oses]