Distributed Learning over Massive XML Documents in ELM Feature Space

Algorithm 2

Reducer of DXRC.

 Input: term, list(docID, element, times, sum

 Output: training samples matrix in the form of position, tfidf

(1) Initiate HashMap mapDocEleTF;

(2) Initiate HashMap mapTDocs;

(3) = DistributedCache.get(“totalDocsNum”);

(4) weights = DistributedCache.get(“elementWeightsVector”);

(5) foreach     do

(6)  weightedDocEleTF = weightsdocId, element  * itr.times/itr.sum;

(7)  mapDocEleTF.put(docId, element, weightedDocEleTF);

(8)  if    then

(9)    newTimes = mapTF.get(docId) + irt.getValue().times;

(10)    mapTDocs.put(docID, newTimes);

(11)  else

(12)    mapTDocs.put(docID, itr.getValue().times);

(13) docsNumber = mapTDocs.size();

(14) idf = ;

(15) foreach     do

(16)  position = itrDocEleTF.getKey();

(17)  tfidf = itrDocEleTF.getValue() idf;

(18)  emit(position, tfidf);