web directories

how are the big directories relating their categories

[topr8]

am developing a largish database of products for a site.

after a fair aount of research i've concluded that the products are going to be categorized in a similiar way to the way the odp categorises its products (in their case the products are external links of course)

the idea is that the visitor will click on a category and see the products in that category (subcategory,subsubcategory .... whatever) and will also be presented with links to other "related" categories.

naturally it is easy enough to show the categories below or immediately above a given category but i'm more interested in what i'd call "sideways links" ...

that is when you drill down at odp you will get the normal subcats listed but also cats/subcats like

anothercat@

i'm wondering if anyone knows if somekind of algorithm is used to calculate these related categories or if they have been hand coded into the db, eg has someone actually gone around making relationships within the db or what?

any ideas

[mack]

The @links are to categories that are similar. I dont know for sure how the ODP do this but a lot of directory software systems allow you this option.

Im most cases when you create a category you will have an option to designate similar categories.

Hope this is of some help.

Mack.

[topr8]

thanks, figured the relating was done by hand which was the direction i was going, fine for this project as there are only about 500 categories,

but i was hoping to develop the idea further for another project and employ some kind of automation