Tweak rank/relevance algorithm

[pazpar2-moved-to-github.git] / doc / pazpar2_conf.xml

diff --git a/doc/pazpar2_conf.xml b/doc/pazpar2_conf.xml
index 6eb2032..454cc02 100644 (file)
--- a/doc/pazpar2_conf.xml
+++ b/doc/pazpar2_conf.xml
@@ -262,17 +262,33 @@
             Specifies that this element is to be used to
             help rank 
             records against the user's query (when ranking is
-            requested). The value is an integer, used as a
+            requested). 
+            The valus is of the form 
+            <literallayout>
+             M [F N]
+            </literallayout>
+            where M is an integer, used as a
             multiplier against the basic TF*IDF score. A value of
-            1 is the base, higher values give additional
-            weight to 
+            1 is the base, higher values give additional weight to
             elements of this type. The default is '0', which
             excludes this element from the rank calculation.
            </para>
            <para>
+            F is a CCL field and N is the multipler for terms
+            that matches those part of the CCL field in search.
+            The F+N combo allows the system to use a different
+            multipler for a certain field. For example, a rank value of
+            "<literal>1 au 3</literal>" gives a multipler of 3 for
+            all terms part of the au(thor) terms and 1 for everything else.
+           </para>
+           <para>
             For Pazpar2 1.6.13 and later, the rank may also defined
             "per-document", by the normalization stylesheet. 
            </para>
+           <para>
+            The per field rank was introduced in Pazpar2 1.6.15. Earlier
+            releases only allowed a rank value M (simple integer).
+           </para>
           </listitem>
          </varlistentry>
          
@@ -547,7 +563,41 @@
         </para>
        </listitem>
        </varlistentry>
-       
+
+       <varlistentry>
+       <term>ccldirective</term>
+       <listitem>
+        <para>
+         Customizes the CCL parsing (interpretation of query parameter
+         in search).
+         The name and value of the CCL directive is gigen by attributes
+         'name' and 'value' respectively. Refer to possible list of names
+         in the
+         <ulink 
+             url="http://www.indexdata.com/yaz/doc/tools.html#ccl.directives.table">
+          YAZ manual
+          </ulink>.
+        </para>
+       </listitem>
+       </varlistentry>
+
+       <varlistentry>
+       <term>rank</term>
+       <listitem>
+        <para>
+         Customizes the ranking (relevance) algorithm.
+         Attribute 'cluster' is a boolean
+         that controls whether Pazpar2 should boost ranking for merged
+         records. Is 'yes' by default. A value of 'no' will make
+         Pazpar2 avergage ranking of each record in a cluster.
+        </para>
+        <para>
+         This configuration was added in pazpar2 1.6.18.
+        </para>
+       </listitem>
+       </varlistentry>
+
+<!--       
        <varlistentry>
        <term>set</term>
        <listitem>
@@ -559,7 +609,7 @@
         </para>
        </listitem>
        </varlistentry>
-       
+-->   
        <varlistentry>
        <term>settings</term>
        <listitem>