A Method for Classification based on Association Rules using Ontology in Web Data

This paper shows a new method based on association rule mining and ontology for the classification of web pages. This work is pruning of association rules, generated by mining process. The main complexity arises due to the fact that there are various number of text documents that are considered for generating the association rules using the A-priori algorithm. But these rules that were generated are not based on the semantic knowledge. In order to obtain the most accurate rules we gone for the construction of the ontology, based on the domain knowledge. With this domain knowledge we design various operators which are helpful in reducing the rules generated. Thus the various rules that we get are semantically correct with regards to the domain selected. We use the high confidence value based classifier for classifying the given text document to that particular domain. Association rules are mined from this matrix using A-priori algorithm. Based on the high confidence value, a new text document is classified into one of the predefined classes. In general, from association rule mining, a huge amount of association rules are mined. All the association rules generated may not be useful for the classification purpose. So, In order to reduce the irrelevant association rules, we need semantic knowledge. For this purpose, propose new domain specific ontology to overcome this drawback of association rule mining method.