The need of effective and efficient method to retrieving non-Web-enabled and Web-enabled information entities is essential, due to the fact of inaccuracy of the existing search engines that still use traditional term-based indexing for text documents and annotation text for images, audio and video files. Previous works showed that incorporating the knowledge in the form of concepts into an information retrieval system may increase the effectiveness of the retrieving method. Unfortunately, most of the works that implemented the concept-based information retrieval system still focused on one information format. This paper proposes a multi-format (text, image, video and, audio) concept-based information retrieval method for Cloud environment. The proposed method is implemented in a laboratory-scale heterogeneous cloud environment using Eucalyptus middleware.  755 multi-format information is experimented and the performance of the proposed method is measured.