Posted by narres on April 26, 2006 at 3:04pm
Should have automated:
- Related content (depending on score, age; should be definable)
- Keywords generation
- View related external data (dmoz, wikipedia)
- Category assignment
- Language detection
Maybe importer of dmoz, wikipedia data like:
http://www.seitenplatz.de/Top (simple example)
It's related to modules spam. nodewords, whatsrelated, search

Comments
Yahoo! also has a great keywords API
it could be used for the automatic assignation of keywords. Can someone help write this up a bit?
.
Automatic/Manual generation of interwiki like urls?
Related (external) data
If this would be an article about Drupal with some text to be analysed
(...)
Related
Articles:
http://groups.drupal.org/node/156
(...)
Websites:
http://www.drupal.de/
(...)
WikiPedia:
http://de.wikipedia.org/wiki/Drupal
http://de.wikipedia.org/wiki/Content_Management_System
http://de.wikipedia.org/wiki/Framework
I wrote a "Similar Entries"
I wrote a "Similar Entries" module that uses MySQL fulltext-indexing to come up with a set of related articles. Deekayen now manages the project.
Yes I know, but first I have
Yes I know, but first I have to beg for pardon for the delay of this answer (currently I'm working in a custumers office and there is only IE allowed. BTW: This site doesn't work with IE ;)).
Now the details:
I had experienced a bit with "Similar Entries" on sites like http://pressemitteilung.ws/node/view/9218, but what it does is a fulltext search, so I implemented a template-based prototypw wich uses the search-table generatet from search.module and some worlist tables.
The result is a much more exact score and performance.
But this is not a well working generic solution, cause it's hard coded and specialized.
The context-suite should be able to build a well working context based on tables like wordlist, wordblacklist, wordwhitelist or something like that, where wordlist should be as small as possible, but big as nessesary.
Sorry, that I've forgot to list your module.
Thomas
great idea
This sounds like a great idea