Exploiting Semantic Annotations and -Learning for Constructing an Efficient Hierarchy/Graph Texts Organization

Algorithm 1

TRN algorithm.

Input: List of Sentences of The document .

Output: List_DSO /^*list of DSO objects and its MFs with attributes. /^*

Procedure

{   ← New document

 List_DSO ← Empty list {list of sentences object of document};

 for  each sentence   in document  do

  AlchemyAPI(); /^* This function calls AlchemyAPI to determine all MF and its NER. /^*

  for  each feature  ∈ in DSO do

   get_MF_Attributes(); /^* This function determines attributes of each MF. /^*

   OpenNLP (si); /^* This function calls OpenNLP to determine POS and NER for MF and attributes. /^*

   compute_MF_score(); /^* This function determines the score of each MF. /^*

  end

  List_DSO.add(DSO);

end}