Posted by Jerry Stuckle on 11/08/07 21:57

Animesh K wrote:
> Hello All:
>
> While the "related-pages" idea is common, I wasn't sure how to implement
> it. Here is the problem:
>
> 1) Let's say my website has 100 or 1000 or 10000 articles.
> 2) Each article has bunch of keywords (e.g., "php, mysql, database,
> navigation")
> 3) Two articles j and k have l(j, k) common keywords.
> 4) Each article j should display a few related articles, drawn randomly
> from articles having common keywords.
> 5) The method should be robust to new entries (in keywords, as well as
> articles).
>
> I wonder if there is an efficient solution around this problem.
>
> Best regards,
> Animesh
>

That's not easy. Are you keeping the articles in a database or text
files? If the former, you can search the database.

But the real problem here is defining what your keywords are. And
keywords might not be enough. For instance, if you have an article on
flowers, 'rose' might be a keyword. But you wouldn't call an article
about how the 'space shuttle rose into the sky' as related. So an
automated solution isn't great.

Maybe just scan the articles as they come in and pick out certain
keywords, similar to metatags in html. Then build a database of keywords.

Or, do it automatically and come back and modify the list manually.

--
==================
Remove the "x" from my email address
Jerry Stuckle
JDS Computer Training Corp.
jstucklex@attglobal.net
==================

[Back to original message]