If you want to query your XML-file, XPath is the way:

• The offical specification
• A tutorial
• The Perl-module: XML::XPath

And here is an example (I slightly changed your XML-file so all elements have different content):

use strict;
use XML::XPath;
use XML::XPath::XMLParser;

my $xp = XML::XPath->new( xml => *DATA );

print 'There are ', $xp->find('queries/query')->size, " 'query' nodes.
+\n";
foreach my $element (qw/topN layer filter name datatype/) {
    print "$element: ",
        (
        join ', ',
        map { XML::XPath::XMLParser::as_string($_) =~ />([^<]*)</ }
            $xp->find("queries/query/$element")->get_nodelist
        ),
        "\n";
} ## end foreach my $element (qw/topN layer filter name datatype/)

__DATA__
<?xml version='1.0'?>
<queries>
    <query>
        <name>topHosts 20</name>
        <layer>LINK 20</layer>
        <topN>20</topN>
        <datatype>topHosts 20</datatype>
        <filter></filter>
    </query>
    <query>
        <name>topHosts 120</name>
        <layer>LINK 120</layer>
        <topN>120</topN>
        <datatype>topHosts 120</datatype>
        <filter></filter>
    </query>
</queries>
[download]

Output:

There are 2 'query' nodes.
topN: 20, 120
layer: LINK 20, LINK 120
filter: 
name: topHosts 20, topHosts 120
datatype: topHosts 20, topHosts 120
[download]

And yes I know that the regex to extract the text-data from the nodes is very crude ...

Update: added a code example.

Update2: simplified and shortened code