Hi Sharma,

I have described my approach in

Chapter Guide for Pragmatical Modelling of Ontologies in Corporate Settings

Hope it is helpful,

Thomas

Hi Chandan

Domain specific Ontology can be built by extraction of important terms, concepts, attributes and relations from unstructured text(documents) using machine learning techniques like text classification with discriminative features selection algorithms and natural language processing with Part-Of-Speech taggers (POS taggers) and parsers. Enclosed publication has a detailed explanation of these.

Article A survey of ontology learning techniques and applications

Hi Chandan Sharma

ontology learning is certainly an option but you I would preliminarily consider (automated) methods matching docs to existing ontologies for the given domain instead of introducing new names/defs. for existing classes and properties.

You might start with the generic ontology in the Reuters Corpora: https://trec.nist.gov/data/reuters/reuters.html