In global e-commerce, semantic document exchange ensures unambiguity and shares same meaning for documents sender and receiver cross different natural languages (e.g., English to Chinese), as the difference makes the translation between cross-language documents complex and inaccurate. This paper proposed a novel collaborative framework of semantic document exchange through a mediation approach which provides a sentence-based machine natural language (MNL). Here, the MNL constructs a grammatical case structure to compatible with all languages. Thus, the mediation framework enables the sentence computer-readable and understandable without ambiguity, simultaneously the MNL can improve accuracy and reduce the complexity of mapping in semantic document exchange.