Algorithm of matching webpage and multi-domain ontology

A problem for searching webpage based on multi-domain ontology is how to precisely judge the relevance between webpage and ontology. This paper proposes a new algorithm based on Floyd algorithm for judging relevance. Firstly, making use of rough set theory to extract attributes from webpage. Then applying the HowNet to find out the synonyms set of the core. Next finding out common keywords and computing the minimal path and the weighting minimal path between the common keywords and ontology. Lastly, deciding the belonging of webpage through comparing the result with the threshold. An experiment is showed to prove this algorithm to be available.