G'day sunil9009,

I strongly suspect poorly named variables are one of your biggest problems here:

• %hash — we already know it's a hash (by the % sigil), beyond that the name is meaningless: it neither reflects the keys (which are customer names) nor the values (which are times).
• $word1, $word2, $word3 — all meaningless and tells us nothing about the data they hold.
• $user — started off as customer names, now they're users.
• $loc — fairly common as an abbreviation for "location" except it holds customer names not locations.

Not surprisingly, when you got to printing your results, you had no idea what data was held where!

You probably want something closer to this:

$ perl -Mstrict -Mwarnings -le '
    my @file_data = (
        "12:10 a  america",
        "12:11 b  bombay",
        "12:12 c  calcutta",
        "12:13 a  australia",
        "2:30   b  bhutan",
        "3:40  n    neterland"
    );

    my %name_data;

    for (@file_data) {
        my ($time, $name, $loc) = split;
        push @{$name_data{$name}{time}} => $time;
        push @{$name_data{$name}{loc}} => $loc;
    }

    for my $name (sort keys %name_data) {
        print "$name: @{$name_data{$name}{time}} : @{$name_data{$name}
+{loc}}";
    }
'
a: 12:10 12:13 : america australia
b: 12:11 2:30 : bombay bhutan
c: 12:12 : calcutta
n: 3:40 : neterland
[download]

Other issues:

• You haven't used strict.
• You haven't used warnings.
• You've used the 2-argument form of open.
• You call die without a message. If you don't want to hand-craft a message, use autodie instead.