Suggest keyword based page relevance algorithm

[hsarkar]

Hi,

I have to use a page relevance algorithm in one of my projects, but am not able to come out with an efficient one.

In my database I have number of articles which is a combination of Title and Body, I also have meta keywords associated with the article, then I have pre-processed a list of regular keywords(with frequency of occurance) and stem keywords(again with frequency), I also have the position of each keyword in the article.

I know that if a search keyword appears in Title / Meta it will have a high rank, then if the keywords appear close to each other again the page ranks higher, last consideration should be frequency of occurance of keywords in the body of text.

Can anyone suggest an algorithm which takes all the above considerations and returns a float in the rance {1, 10} which is the rank of the page.