New semantic indexing and search system based on ontology

Information retrieval becomes a very complex process for search engines on the Web, this is due to, first, the staggering growth speed of the number of web site and, in the other hand, the search algorithms by keywords (terms) used currently are not suitable to better exploit this huge information quantity. These elements make the information retrieval by the current search engines very difficult and does not meet the users' needs. In this paper, we propose a new method of semantic information retrieval based on ontology. Our method allows the indexing and searching engine to take in consideration the documents semantic level, which significantly improves the quality of search results. We have implemented our approach on an indexing and searching engine based on ontology (called MIRO Moteur d'Indexation et de Recherche base sur les Ontologies). MIRO offers a multilingual semantic search of documents using concept instead of term. Additionally, MIRO offers a guided search tool, and a tool for an automatic enrichment of ontology. Moreover a comparison of results between MIRO and PhpDig (an open source search engine) is presented.