It would help if you showed us what you've already tried, if anything. The basic structure of the program is fairly straightforward, but the actual implementation can vary, depending upon how the input data looks and how you need output. Does your input data really begin with ","lliu5@yahoo.com","? If so, that's a strange beginning to the line. In any event, here's one structure for doing this (pseudo-code):

foreach line of input:
    split line into tokens
    foreach token in tokens
        if token is valid email and not from "domain.com"
            add token to hash
        end if
    end foreach token
end foreach line

A straight-forward implementation would go something like this (untested):

use Text::CSV_XS;
use Email::Valid;
my %email;
my $csv = Text::CSV_XS->new;

while (my $line = <DATA_FILE>) {
    chomp $line;
    if (my $status = $csv->parse($line)) {
        my @columns = $csv->fields;
        foreach my $column (@columns) {
             $email{$column} = 1 if good_email($column);
        }
    }
    else {
        warn "Could not parse ",$csv->error_input,"\n";
    }
}

sub good_email {
    my $email = shift;
    if (Email::Valid->address($email) && $email !~ /domain\.com$/i) {
        return $email;
    }
    return;
}
[download]

When that's done, your email hash keys have all unique emails. However, the list is unsorted and I my domain check is a bit dodgy, but this does give you an idea of how to proceed.

Unlike Coruscate's solution, this is fairly procedural, but it's also very easy to read.

Cheers,
Ovid

New address of my CGI Course.
Silence is Evil (feel free to copy and distribute widely - note copyright text)