You have records with two fields seperated by colon. You can just split on that to get the arrangement you want. Putting that in a hash is attractive, but doesn't preserve order, so lets make it an array of arrays instead. You probably want to trim leading and trailing whitespace, too.

for (@DATA) {   # this transforms @DATA, destructive
    $_ = [map {/^\s*(.*?)\s*$/} split ':', $_, 2];
}
[download]

The split on colon does the field seperation, then the mapped match trims leading and trailing space. Square brackets make the result an array reference.

To print the new @DATA in the form you want,

for (@DATA) {
    printf "FIRST TERM - %s\nSECOND TERM - %s\n", @$_;
}
[download]

Update: holli++, I did miss that. We do see what we expect, don't we? Second try,

for (@DATA) {
    $_ = join ' ', split;    # normalize whitespace
    $_ = [split /[:\s]/, $_, 2];
}
[download]

That's simpler to read, too.