Linking text and visual concepts semantically for cross modal multimedia search