The "semantic web" is supported by ontologies and background knowledge. For several reasons, it is unlikely that his formal knowledge will be encoded in the very same language. Taking advantage of the semantic web requires to be able to compare, transform and compose these formal knowledge languages. The main purpose of this paper is to study the problem of language interoperability. In this paper, we present a new approach based on concept lattice. It is a branch of the lattice theory that allows us to identify meaningful groupings of elements that have common properties with the aid of formal concept analysis (FCA). The established lattice is a set of knowledge representation Languages partially ordered with regard to the transformability from one language to another by preserving a particular property. This approach is more general than the existing approaches and convenient. For the same set of languages, there can be several lattice structures depending on the property for transforming languages. Here, We discussed semantic property transformations of different strength. These properties involve syntax, interpretation, consequence and consistency preservation.