CC BY 4.0Stepanov, Vadim K.2025-06-102025-06-102025-06-09https://youtu.be/EOReF_OsvQE?si=YkwWcADuM-XOWq4u&t=2664https://repository.ifla.org/handle/20.500.14598/4057Bibliographic Activity in Social Sciences in the Era of Neural Networks and API Interfaces by Vadim K. Stepanov (Moscow, Russia) The total digitalization of information transfer processes and the parallel development of several standards and technologies at once have formed a new paradigm of bibliographic and, in general, information services for the scientific sphere. The foundation of the transformation was the adoption of international standards, DOI and ORCID. The method of providing scientific papers has also been unified: in most global scientific journals, when the author uploads a manuscript, along with the full text, all accompanying metadata are entered, launching the entire further cycle of bibliographic information. When a scientific paper is published, information about it is automatically transferred to Crossref (the global DOI registration agency), which has become a giant global repository of bibliographic information continuously received from publishers. Crossref is an open resource, information from which can be obtained by anyone without restrictions. API technology is used to borrow data from the Internet, which has become the basis for the bibliographic data exchange system: due to API, bibliographic information is continuously transferred from resource to resource in previously unimaginable gigantic volumes in the background without any human intervention. The opportunities that have opened up have been taken advantage of by companies that can be described as global discovery services or global bibliographic platforms. Their goal is to unite the entire world's flow of scientific publications to provide a full cycle of information services to any user in strict accordance with their individual information needs. Receiving bibliographic information from many publishers and specialized services, they have already accumulated arrays of hundreds of millions of records. The data is subject to intellectual processing – information obtained from different sources is combined into a single bibliographic record. Based on this array, users are provided with all traditional types of information support for research activities in a fully automated mode. Today, such companies include Semantic Scholar, The Lens, OpenAlex, Scilit, Google Academy, ResearchGate, Scopus, and Web of Science. In Russia, the equivalent of such an academic service is Elibrary. In the future, information support will increasingly shift towards global bibliographic platforms that provide comprehensive data with maximum comfort. Bibliographic services around the world need to take this fact into account and shift the focus to processing printed retrospective data and the flow of serials that are not classified as scientific publications.enhttps://creativecommons.org/licenses/by/4.0/Social sciencesArtificial intelligenceResearch methodsEvents MaterialVadim K. Stepanov