Relevant Pages in semantic Web Search Engines using Ontology

In general, search engines are the most popular means of searching any kind of information from the Internet. Generally, keywords are given to the search engine and the Web database returns the documents containing specified keywords. In many situations, irrelevant results are given as results to the user query since different keywords are used in different forms in various documents. The development of the next generation Web, Semantic Web, will change this situation. This paper proposes a prototype of relation-based search engine which ranks the page according to the user query and on annotated results. Page sub graph is computed for each annotated page in the result set by generating all possible combinations for the relation in the sub graph. A relevance score is computed for each annotated page using a probability measure. A relation based ranking model is used which displays the pages in the final result set according to their relevance score. This ranking is provided by considering keyword-concept associations. Thus, the final result set contains pages in the order of their constrained relevant scores.