Building a Recommendation Engine: The XELOPES Library - Realtime Data Mining - page 241

Example 12.3
// Create category hierarchy:
CategoryHierarchy cah ¼ new CategoryHierarchy();
// Parent category for sharp cutlery:
Category catSharp ¼ new Category("sharp");

// Relations:
cah.addRelationship(catSharp, catKnife); // knife is sharp
cah.addRelationship(catSharp, catFork); // fork is sharp

// Assign hierarchy to cutlery:
cutlery.setTaxonomy(cah);

■

We conclude with an example of MiningDataSpecification.

Example 12.4
// Create object of metadata 'meal':
MiningDataSpecification meal ¼ new MiningDataSpecification
("meal");

// Create numeric attribute 'calories' and add to metadata:
NumericAttribute

¼

calories new NumericAttribute

("calories");
meal.addMiningAttribute( calories );

// Create numeric attribute 'numberOfGuests' and add to
metadata:
NumericAttribute numberOfGuests ¼ new NumericAttribute();
numberOfGuests.setName( "number of guests" );
meal.addMiningAttribute( numberOfGuests );

// Add previous categorical attribute 'cutlery' to metadata:
meal.addMiningAttribute( cutlery ); ■

12.1.2.2 The Coordinates: MiningVector

After we have modeled the basis of the attribute space by MiningDataSpecification ,
we will now model the coordinates of a vector. This is done through the class
MiningVector .
MiningVector contains a reference metaData (of the class MiningDataSpeci-
fication ) to its basis and an array of real values which stores the coordinates of the
vector. MiningVector extends the CWM class Object of the CWM resource package
Instance since it represents an instance of the data described by MiningDataSpeci-
fication (Fig. 12.6 ) .