Research Article

Gamma-Poisson Distribution Model for Text Categorization

Algorithm 1

Algorithm of classifier using gamma-Poisson modeling. In the training phase, the procedure of learning over (entire training documents) is given, while in the test phase, the procedure to classify one test document is described. is the set of training documents belonging to a class and is a document vector in . Note that , and thus the time complexity of the training phase is estimated as . For the test phase, the complexity is found to be because in line 02 of the pseudo code is calculated through the entire summation over (see (13)).
gamma-Poisson classifier for training phase
01 for each
02  for each
03   for each
04     count tokens of term in
05    
06    
07   end for( )
08   
09   
10   
11    rational Approximation
12   
13  end for( )
14 end for( )
gamma-Poisson classifier for test phase
01 for each
02  score
03 end for( )
04 return   score