The Spoken Slovene portal is part of the Large Language Models for Digital Humanities (LLM4DH) project (2024–2027), which focuses on research into the development of large language models for Slovene and their use in digital humanities. The portal can also be used to submit recordings collected in collaboration with other research projects. The expansion of the portal to include accessible content for the general public and additional features is funded through the Spoken Slovene 2.0 project, supported by the Ministry of Culture of the Republic of Slovenia from 1 May 2026 to 31 October 2027.
Previous speech collections
ROG
A collection of recordings of public and private speech with transcripts and manually reviewed annotations of words, grammar, disfluencies and utterance meaning.
Purpose:
Training machine-learning models for speech annotation and linguistic analysis.
Projects:
A Treebank-Based Approach to Research on Spoken Slovene – SPOT (2022–2024) | Basic Research for the Development of Spoken Language Resources and Speech Technologies for the Slovenian Language – MEZZANINE (2022–2025)
Artur
A large collection of more than 1,000 hours of speech from various sources, including more than 800 hours with accurate manual transcripts.
Purpose:
Development of automatic speech recognition for Slovene and linguistic research.
Projects:
Development of Slovene in a Digital Environment – RSDO (2020–2023).
Gos
A 100-hour collection of speech in public and private contexts, with manually produced transcripts and automatically added linguistic annotations such as part of speech, gender, case and number.
Purpose:
A source of information on Slovene vocabulary, the grammatical properties of words, their phonetic form and pronunciation, supporting more comprehensive dictionary and grammatical descriptions of Slovene.
Projects:
Communication in Slovene – SSJ (2008–2011).
Portal Content
Dr Darinka Verdonik, Associate Professor of Slovene Language, research in spoken Slovene and the development of speech resources, University of Maribor, Faculty of Electrical Engineering and Computer Science
Andreja Bizjak, MA, Teaching Assistant in Media Communications, research in speech resources, University of Maribor, Faculty of Electrical Engineering and Computer Science
Technical implementation
Dr Gregor Donaj, Assistant Professor of Electrical Engineering and researcher in language technologies, UM FERI
Design and illustrations
Samo Kramberger
Neli Kramberger
Distinguished citizen scientists
Dominik Ivič
Legal documents
Boštjan Makarovič, Cristina Contero Almagro, Aphaia
Custodian of the recordings
Institute of Electronics and Telecommunications, Faculty of Electrical Engineering and Computer Science, University of Maribor, govorjena-slo@um.si
Data Controller
University of Maribor, Slomškov trg 15, 2000 Maribor, +386 (0) 2 23 55 280, rektorat@um.si
Data Protection Officer
Assoc. Prof. Dr Miha Dvojmoč, Faculty of Criminal Justice and Security, Kotnikova 8, 1000 Ljubljana, dpo@um.si