thanks,Renoir

That would be great.

I didn't think I would get as much of a response to this.
i have spent a lot of work on this project. previously the search engine was written using PHDBase and I was asked to re-write the search engine as PHDBase had a few downfalls.

Just to explain myself further. I work for a company that has a large list of names and addresses (3.7Million). the task in hand is to return the nearest possible matching person from search criteria entered namely

First Name
Surname
House No
Street
Town
And a choice of two counties

The user can enter any amount of Search information they require ( if they are not sure which county the required person is in they can select two counties)

the main objective of this search engine is speed !!.
the result has to return an answer in less than 10 Seconds. Which I am happy it is doing it, However I know I can speed it up

And this is how
If the user entered
Joe Bloggs 1 Main Street
I would first of all retrieve all of the exact matches to Joe Bloggs and build a list of Name and Addresses.
I would then traverse this list of Addresses and assign a matching percentage Figure to the address

So effectively what I need to do
is search my potential list of addresses for the occurrance of the specified address.
by the way, I have incorporated soundex search on the name, but this is only used as a last resort if the name is not found !!

hope this explains things

Mark








>>Hi All
>>
>>i was wondering, does anyone know where one could get a utility that compares two strings (a+b) and returns a percentage match
>>
>>Eg A = "JOE BLOGGS HOUSE"
>> B = "My Friend Joe Bloggs Lives In a huge House Down the Road"
>>
>>The utility would return a percentage figure of a match say 80% as it found "Joe Bloggs" and "House" in the sentence
>>
>>
>>Thanks
>>Mark
>
>Mark,
>
>Interesting... You would be looking for the percentage of matching words, right? I know that Nadya has a word parser as part of a larger process she has posted. I'll look in my libraries and see if I can find it. With that you could use something like AT to determine the matches in the strings. I might be slow though depending on the volume of what you're doing. There also may be an easier way...
>
>Renoir