During the recent years the Web has been developed rapidly making the efficient searching of information difficult and time-consuming. In this work, we propose a web search personalization methodology by coupling data mining techniques with the underlying semantics of the web content. To this purpose, we exploit reference ontologies that emerge from web catalogs (such as ODP), which can scale to the growth of the web. Our methodology uses ontologies to provide the semantic profiling of users’ interests based on the implicit logging of their behavior and the on-the-fly semantic analysis and annotation of the web results summaries.