Fuzzy Ontology Based Web Mining Approach for Extraction of Semantic Web Documents

Ontology represents relationships among set of terms and concepts in hierarchical fashion. Ontology plays crucial role in formulization of information related to given domain. Understanding these ontologies without having sufficient knowledge of ontology editors is like working on project without knowing its requirements. Traditional text mining methods and aero-text systems for extracting key phrases have been used but it needs to be improved to support large scale ontology constitution for real world applications. An ample amount of documents present on web puts the users in state of dilemma. Relevance means how closely the given query matches large number of documents. The paper proposed fuzzy ontology based approach that retrieves information from web documents by using fuzzy relations and semantic context vectors. It discovers fuzzy ontology rather than textual descriptive ontology with crisp features only. The output membership fuzzy functions are produced by simulation tool named as MATLAB. The validation of proposed approach is done by evaluating information retrieval performance in two specific domainsweather domain (web pages containing information about weather forecasting and analysis) and Google TM collection (web pages containing news).