Internet usage is increasing at a rapid rate. Anyone can access all sorts of information from the web at any time. But the language diversity of web pages restricts this access of information. Cross Lingual information Retrieval provides the solution to this problem, by allowing user to post their query in the native language and retrieve the documents in other languages. Machine readable dictionaries are the most economical resource of cross language translation. These dictionaries usually provide more than one translations in target language for source query terms. These translations need to be disambiguated to achieve the best translation for a query word. Once we achieve the correct translations, the cross lingual information system can match the monolingual information retrieval on performance grade. In this paper we propose a framework for query translation and disambiguation to achieve an efficient cross lingual information system.