You could try one of the /sounds like/ type of modules from CPAN. Text::Metaphone and Text::Soundex are two that I have used. Both have limitations. It really depends on what you are willing to consider a "close match". Do you want 'Don Banks' to be a close match for 'Dawn Binx'? You may get odd results.

Alternately, you will need to better define what constitutes a "close match". Is it something that a human would recognize, but is hard to define in programmatic ways? Off-by-one type matches you may be able to program, but it could be error-prone and expensive (time intensive to run).

Any additional info you can supply here would help us better guage what you are trying to accomplish.

Hi, Thanks everyone, I looked at the soundex module, but somewhere it said that it only considered English words and pronounciations... some of the names I have to work with are Indian, Chinese and even African :-( The problem in more details: I have a group of people calling two different helpdesks. They give their names to be identified. I need to find out which people call both helpdesks, and I can only use their names as identification: they do not give their IDs, and their departments change quite often so that would be not useful to compare. But these can be spelled wrongly by the helpdesk personnel; also, for example, a Robert Carlos could give his name as Bob Carlos sometimes :-( I don't know that I can do much about short forms of names, but I would like different spellings to be identified as the same person as much as possible. Thank you very much, Eyi.

Lingua::EN::MatchNames would be a start - I'm not sure how much support there is for indian/chinese/african names

Thanks Transient. I'm looking at it, but I'm not sure it'd work for non-English names :-( But I'll try and see what results I get.

Thanks ww, this is certainly helpful. Checking it out...

That is to look at a few modules on CPAN, and see if they fit your needs.

• Win32::OLE
• Spreadsheet::ParseExcel
• String::Approx
• Text::Levenshtein

If you need more specific help after that, come back and augment your question with more details.

For "simple" things, I like Spreadsheet::ParseExcel::Simple.

sub offby {
        my($dist,$word,@possibles)=@_;
        my($re)=sprintf(
                q|^%s:%s$|,
                join(q|.?|,map{q|(\w*)|}(0..$dist)),
                join(q|.?|,map{"\\".($_+1)}(0..$dist))
        );
        my(@suggestions);
        #print "user $word regexp $re\n";
        foreach(@possibles) {
                my($str)="$word:$_";
                if($str=~/$re/) {
                        push(@suggestions,$_);
                }
        }
        return(@suggestions)
}
[download]

I assume you have a large sampling of names to work from? Otherwise you could do this by hand, and wouldn't need a script for it.