Precise identification in bibliographic systems is a basic condition to data exchange and carrying out complete and verifiable bibliometric analysis of scholarly output and research capacity of scientific organization or university. Such an analysis can be conducted at different levels - the whole organization level, department, laboratory, research team or individual researcher level. Unique accepted and standardized method for identifying journal articles, authors, their affiliations, and other elements of bibliographic metadata are absent despite the high efforts made in that direction. Initiatives to construct identifiers and systems appear regularly in scientific papers and reports; however, they have not achieved the necessary degree of interoperability. This paper describes the main challenges in identification of various bibliographic metadata and their solutions at the level of scientific organization.