In this phase, the semantic inverted index of a collection
of documents is built. The algorithm of the index
creation starts to manipulate each document of the
collection by extracting and preprocessing its phrases
one after another. The preprocessing operations on the
phrase include the removal of the stop words which are
listed in the stop words list and the stemming. These
preprocessing operations are standard operations in any
information retrieval system. The next operation is the
reasoning of the ontology using the set of words that are
resulted after the phrase preprocessing operation to get a
reference concept from the ontology links between
these words.