A concept-level ontology construction method for automatic summarization

The increased presence of unconventional data sources on the Internet in many fields have led to an exponential increase in the amount of information available. Even though search engines are used to filter the search results, a multitude of unwanted results remain. Automatic summarization has become important in speeding up the information retrieval process. However, automatic summarization usually needs a good ontology, which is not always available for all domains. This paper describes a novel ontology construction approach based on Natural Language Processing (NLP) and Knowledge Representation techniques to facilitate finding important sentences for automatic summarization. This approach is more flexible and less domain dependent than other traditional statistical-based ontology construction methods. We outline an ontology construction that improves the results of automatic summarization.