"Best possible" category for a term?

userlite

Member
Joined
Mar 10, 2008
Messages
2
Hi,

I would like to use DMOZ data - content and structure files to find out the best possible category for a term.

For example: when I enter "bmw" in DMOZ online search, the first result is:

# Recreation: Autos: Makes and Models: BMW
, which is most relevant for term "bmw" - as compared to others such as:

# Recreation: Motorcycles: Makes and Models: BMW (95)
# World: Deutsch: Freizeit: Auto: Marken: BMW (72)
# Business: Automotive: Motorcycles: Makes and Models: Retailers: BMW (31)
# Home: Consumer Information: Automobiles: Purchasing: By Make: BMW (11)

How can I achieve this using the rdf files? Out of all the possible categories, how do I pick the right one automatically?

Any help is appreciated. Thanks.
 

chaos127

Curlie Admin
Joined
Nov 13, 2003
Messages
1,344
I guess you would need to design a search algorithm that looks at the contents of each category (category name, path, titles of listed sites, descriptions of listed sites) and gives is a score based on occurrences of the search term(s) entered. Then you just return the category with the highest score. That's pretty much what the search function on http://www.dmoz.org/ does to come up with the category choices.
 
This site has been archived and is no longer accepting new content.
Top