Changes between Version 7 and Version 8 of similarity_SimFlood

Timestamp:

09/14/06 17:11:56 (18 years ago)

Author:

lqu

Comment:

--

similarity_SimFlood

@@ -54 +54 @@
 Alone with formula basic and inria benchmarks. precision 0.68, recall 0.31.
 
-=== Specification ===
+=== Suggested configuration ===
+If there is a alignment from another similarity measure as input, fixpoint computing formula C and growth policy All together is a good combination. 
+One alternative is to use growth policy SinglePCG for faster computation, because it generates much less PCGs. 
+If there is a no alignment as input(pure structure comparision). One possibility is to use fixpoint computing formula Basic and growth policy. Sometimes better results can be achieved if SF_OPTIMIZED_BIND_INSTANCE is set to be true, growth policy is PCGStepwise and SF_BIND_WHOLE_DATA_TAXONOMY is set to be false
 ==== Intitialisation ====
 The SimilarityMeasure class is
@@ -64 +67 @@
 || '''Parameter name''' || '''ValueType''' || '''Default''' || '''Description''' ||
 || PARAM_TAXONOMY_ONLY || Boolean || {{{FALSE}}} || Defines whether the properties should be included in the connectivity graph. Thus, this option is vital if you want to calculate similarities between properties as well.||
+|| SF_OPTIMIZED_BIND_INSTANCE || Boolean || {{{FALSE}}} || Defines whether the instances should be included in the connectivity graph.||
+|| SF_OPTIMIZED_FORMULA || SimilarityMatcher.Formula || Formula.Basic || fixpoint computing formula according to the paper of Similarity Flooding.||
+|| SF_OPTIMIZED_GROWTH_POLICY || SimilarityMatcher.GrowthPolicy || GrowthPolicy.All || The way to build paarwise connectivity graph.||
+|| SF_PROPAGATION_COEFFICIENTS || SimilarityMatcher.PC || PC.inv_prod || propagation coefficients, details see melnik paper.||
+|| SF_BIND_WHOLE_DATA_TAXONOMY || Boolean || {{{TRUE}}} || if bind all the supported datatypes into the PCGs||
+