Item2486: Daniel Oderbolz added logarithmic normalization of term wei…

…ghts, making it very nice tag clouds git-svn-id: http://svn.foswiki.org/trunk/TagCloudPlugin@5763 0b4bb1d4-4e5a-0410-9cc4-b2b747904278
foswiki · Dec 10, 2009 · ab6062c · ab6062c
1 parent 1453f18
commit ab6062c
Show file tree

Hide file tree

Showing 3 changed files with 24 additions and 1 deletion.
diff --git a/data/System/TagCloudPlugin.txt b/data/System/TagCloudPlugin.txt
@@ -100,6 +100,8 @@ Parameters:
      (default: 10)
    * =lowercase=: switch on/off converting terms to lowercase
      (default: off)
+   * =normalize=: method to normalize the term counts, either logarithmic or linear 
+     (default: log)
    * =plural=: switch on/off counting plurals, if set to off plural terms are matched
      to their singular form (caution: very rudimentary and assuming the input data is all english, 
      use stopwords="on" to get good results)
@@ -191,6 +193,7 @@ Cloud for the above text:
 |  Release: | %$RELEASE% |
 |  Version: | %$VERSION% |
 |  Change History: | &nbsp; |
+|  10 Dec 2009: | added logarithmic normalization (Foswiki:Main/DanielOderbolz) |
 |  25 Nov 2009: | fixed grouping in case sensitive tag clouds |
 |  03 Sep 2009: | made sorting case insensitive; added case sensitive extra param |
 |  25 Aug 2009: | added custom fields for terms in the tagcloud (Foswiki:Main.OliverKrueger) |

diff --git a/lib/Foswiki/Plugins/TagCloudPlugin.pm b/lib/Foswiki/Plugins/TagCloudPlugin.pm
@@ -24,7 +24,7 @@ use vars qw(
 );
 
 $VERSION = '$Rev$';
-$RELEASE = 'v2.11';
+$RELEASE = 'v2.20';
 $NO_PREFS_IN_TOPIC = 1;
 $SHORTDESCRIPTION = 'Renders a tag cloud given a list of terms';
 

diff --git a/lib/Foswiki/Plugins/TagCloudPlugin/Core.pm b/lib/Foswiki/Plugins/TagCloudPlugin/Core.pm
@@ -349,6 +349,9 @@ our %stopWords = (
   },
 );
 
+use constant NORMALIZE_LINEAR => 0;
+use constant NORMALIZE_LOG => 1;
+
 ###############################################################################
 sub writeDebug {
   print STDERR '- TagCloudPlugin - '.$_[0]."\n" if DEBUG;
@@ -382,6 +385,7 @@ sub handleTagCloud {
   my $theGroup = $params->{group} || '';
   my $theFilter = $params->{filter} || 'off';
   my $theLimit = $params->{limit};
+  my $theNormalize = $params->{normalize} || 'log' ;
 
   unless (defined($theHeader) || defined($theFooter)) {
     $theHeader = '<div class="tagCloud">';
@@ -403,6 +407,15 @@ sub handleTagCloud {
   $theStopWords = 'off' unless $theStopWords =~ /^(on|off)$/;
   $thePlural = 'on' unless $thePlural =~ /^(on|off)$/;
 
+  if ($theNormalize eq 'linear') {
+    $theNormalize = NORMALIZE_LINEAR;
+  } elsif ($theNormalize =~ /log(arithmic)?/) {
+    $theNormalize = NORMALIZE_LOG;
+  } else {
+    $theNormalize = NORMALIZE_LINEAR; # default
+  }
+
+
   # build class map
   my %classMap = ();
   if ($theMap) {
@@ -447,6 +460,7 @@ sub handleTagCloud {
   my $stopWords; 
   $stopWords = getStopWords($theLanguage) 
     if $theStopWords eq 'on';
+
   foreach my $term (split(/$theSplit/, $theTerms)) {
     $term =~ s/^\s*(.*?)\s*$/$1/o;
     #writeDebug("term=$term");
@@ -488,10 +502,16 @@ sub handleTagCloud {
   my $floor = -1;
   my $ceiling = 0;
   foreach my $term (keys %termCount) {
+
+    # normalization
+    $termCount{$term} = log($termCount{$term}) if $theNormalize == NORMALIZE_LOG;
+
+    # truncation
     if ($termCount{$term} < $theMin) {
       delete $termCount{$term};
       next;
     }
+
     $ceiling = $termCount{$term}
       if $termCount{$term} > $ceiling;
     $floor = $termCount{$term}