The integration is supported by the ontology merging
techniques including measuring similarity, merging classes and
merging properties. Similarity identification is the most
important part in the ontology merging process, which
identifies similar concepts that have to be merged into a single
one [25]. It is been identified that there are two main
approaches for the similarity identification: (1) it is based on
the fact that the ontologies are designed and constructed to
support future semantic integration. That is, the ontologies are
constructed in similar or same structure; and (2) it is based on
heuristics or machines learning techniques that use various
ontology features to discover similar concepts [25]. In this
work, the ontologies meet the first situation. Therefore, the
proposed merging method is based on linguistic analysis of
concepts to calculate the syntactic and semantic similarity
between them.
Overall, the above three phases are guiding the main works
to construct the required domain ontology. Further, the rules of
the above phases will be elaborated in Section as designed
reuse algorithms. Also, there are algorithms designed to
calculate the degrees of reuse the knowledge bases in the
constructed domain ontology and the generated ideas. Thus, the
ontology for the research ideas creation can be constructed
systematically supported by reusing knowledge bases.