In Yewno, the identification of topics that a text deals with is based solely on methods of artificial intelligence and machine learning. Unlike in classical cataloguing systems, they are not associated with a text as a total, but with the text passage in question.
In the test application currently available at the Bayerische Staatsbibliothek, Yewno searches 40 million English-language documents from publications by well-known academic publishing houses such as Cambridge University Press, Oxford University Press, Wiley, Sage and Springer, as well as documents available in Open Access. The search is accessible to the user intuitively: Inputting a keyword or topic, called concept by Yewno, immediately leads to a graphic representation of a semantic network of relevant concepts and their interconnections with regard to content. Thus, it becomes possible to navigate via topical connections up to the located text sources, which are then displayed in the form of so-called snippets.
After the three-month test phase, the feedback by the users will be evaluated first. Whether and when the step from the classical semantic search engine to the semantic discovery service will be taken and what importance applications such as Yewno will acquire within this context is not foreseeable today.
Director General Klaus Ceynowa: "Testing the Yewno software represents an investment in the future to us. We see it as our mission to continuously identify seminal developments and technologies for libraries at an early stage and to test their applicability, capacity for further development and future-proofness."
The software Yewno has been developed by the start-up company of the same name in cooperation with Stanford University, which is also a close cooperation partner of the Bayerische Staatsbibliothek.
Press release for download (in German) (PDF, 60 KB)
Dr. Berthold Gillitzer
Deputy Head of the User Services Department
Phone: +49 89 28638-2659
Press and Public Relations
Telefon: +49 89 28638-2429