This software is provided without any guarantee. The developer, distributor and their organization(s) are NOT responsible for any type of loss caused by this software. 

The directory XML_Input has necessary input files.

adult.xml

This is the input file for the main program. It has following tags

<k>500</k>	This is input value for k-anonymity

<classes>1000</classes>	Number of classes to be suppressed

<writeSuppressDataset>0</writeSuppressDataset>	Flag to write data into weka file after each class suppression

<KullbackLeiblerDivergence>10000</KullbackLeiblerDivergence> Stop the suppression of classes when KullbackLeiblerDivergence becomes more than this input parameter

<NormalisedMutualInformation>10000</NormalisedMutualInformation> Stop the suppression of classes when NormalisedMutualInformation becomes more than this input parameter

<attribute> </attribute>  Describes the name of the attribute


<sensitive>0</sensitive> Describes either the corresponding attribute is sensitive or not

<inputfile>adult_45222.data</inputfile>	Name of the input data file. The input file must be in CSV format

<lines>45222</lines>	Lines to be read from the input file

<namesfile>adult.names</namesfile>	Name of the names file

<outputfile>out.data</outputfile>	Name of the output file. The output will also be generated in CSV format

The data and names file should be in the same folder as of XML input file. The output file is also created in the same folder


The format of the names file is as follows

attributeName	\t hierarchical	

The attribute names and the sequence should be the same as defined in the <attributes> tag of the XML file.

The attribute type (hierarchical) and attributeName is separated by carriage Return key i.e. \t

The hierarchical structure is defined as follows.

{1-100{1-50{1-25}{26-50}}{51-100{51-75}{76-100}}}
					1-100
					|
			----------------------------------------------
			1-50					51-100
			|						|
		----------------			-----------------------
		1-25		26-50			51-75			76-100




The next tags of the XML file deals with the Weka accuracy values

<KInterval>5</KInterval> Defines the interval between the class for the calculation of accuracy values i.e. 1,6,11,16,21..


<J48>1</J48>
<NaiveBayes>1</NaiveBayes>
<Logistics>1</Logistics>
<SVM>1</SVM>

Currently four classifiers are supported and where the value is set to 1 the accuracy values will be calculated for that classifier.

<classifierIndex>8</classifierIndex> Index of the classifier attribute in the .names file

<folds>10</folds> Number of folds during training and test accuracy calculation

<training>90</training> <test>10</test>	Define the ratio between training and test datasets.

<CrossValidationFolds>10</CrossValidationFolds> Folds values during cross validation



 







