In this paper authors perform a research on possibilities of RDF (Resource Description Framework) syntaxes usage for information representation in Semantic Web. It is described why pure XML cannot be effectively used for this purpose, and how RDF framework solves this problem. Information is being represented in a form of a directed graph. RDF is only an abstract formal model for information representation and side tools are required in order to write down that information. Such tools are RDF syntaxes – concrete text or binary formats, which prescribe rules for RDF data serialization. Text-based RDF syntaxes can be developed on the existing format basis (XML, JSON) or can be an RDF-specific – designed from scratch to serve the only purpose – to serialize RDF graphs. Authors briefly describe some of the RDF syntaxes (both XML and non-XML) and compare them in order to identify strengths and weaknesses of each version. Serialization and deserialization speed tests using Jena library are made. The results from both analytical and experimental parts of this research are used to develop the recommendations for RDF syntaxes usage and to design a RDF/XML syntax subset, which is intended to simplify the development and raise compatibility of information serialized with this RDF syntax.