PROJECT TYPE:

• Institutional Scientific-Research Project

PROJECT DURATION:

• 1 October 2025 – 30 September 2029

PROJECT LEADER AND TEAM MEMBERS:

Project Leader:

• Blaženka Martinović, PhD, Full Professor (Faculty of Humanities, University of Juraj Dobrila in Pula)

Project Team Members:

• Mieta Bobanović Dasko, PhD, Assistant Professor (Faculty of Informatics, University of Juraj Dobrila in Pula)
• Nikolina Jurišić, Assistant (Faculty of Humanities, University of Juraj Dobrila in Pula)
• Ratomir Karlović, mag. inf. (Faculty of Informatics, University of Juraj Dobrila in Pula)
• Dario Kukuljan, Technical Associate (University of Juraj Dobrila in Pula)
• Ivan Lorencin, PhD, Assistant Professor (Faculty of Informatics, University of Juraj Dobrila in Pula)
• Mia Rovis, univ. bacc. inf. (Faculty of Informatics, University of Juraj Dobrila in Pula)
• Alma Smajić, mag. inf. (Faculty of Informatics, University of Juraj Dobrila in Pula)
• Nikola Tanković, PhD, Assistant Professor (Faculty of Informatics, University of Juraj Dobrila in Pula)

• Nikola Ljubešić, PhD, Assistant Professor (Knowledge centre for South Slavic languages / Department of Knowledge Technologies, Jožef Stefan Institute)
• Mihaela Matešić, PhD, Full Professor (Faculty of Humanities and Social Sciences, University of Rijeka)
• Elenmari Pletikos Olof, PhD, Associate Professor (Faculty of Humanities and Social Sciences, University of Zagreb)
• Ivan Porupski, Assistant (Knowledge centre for South Slavic languages / Department of Knowledge Technologies, Jožef Stefan Institute) 
• Jelena Vlašić Duić, PhD, Full Professor (Faculty of Humanities and Social Sciences, University of Zagreb)

PROJECT SUMMARY:

The project Croatian Accents Online (CAO) brings together an interdisciplinary team of experts in Croatian linguistics, phonetics, computer science, electrical engineering, and information technology. Its primary aim is to develop innovative digital tools for the learning and teaching of standard Croatian accentuation. The specific objectives of the project include the creation of online applications for practicing and testing the perception and production of Croatian accents (both standard-language and stress-based, as well as regiolectal varieties), the enhancement of these applications through automatic annotation of accent placement, and, ultimately, the development of an audio repository of verbs, which will serve as the foundation for a future online audio dictionary. The tools for practicing accent perception and production facilitate the acquisition of skills essential for the effective use of the audio dictionary and are therefore a prerequisite for the audio integration of Croatian lexical resources, embedded in a communicative context and validated through spoken corpora. The research component encompasses the collection and acoustic processing of a speech corpus (including archetypal speaker pronunciation and regiolectal varieties) and the definition of archetypal pronunciations of accents upon acoustic analyses (of pitch and duration). This will be followed by a comprehensive analysis of the verb corpus, classification into accentual types and subtypes, examination of hypercorrections (based on a parliamentary corpus), and implementation of automatic accent-location annotations. The team from the Faculty of Informatics in Pula, in collaboration with external partners from the Institute in Ljubljana, is responsible for developing the online applications, implementing acoustic analyses, processing accent pronunciation data, and constructing the digital infrastructure (including cloud solutions, the application of large language models, and speech transformers). The project shall contribute to the development of modern educational tools and online resources for the Croatian language, as well as to advancing research in Croatian accentology, phonetics, and methodology for teaching Croatian accentuation. Furthermore, it shall support the digital transformation and enhance the international visibility of the Juraj Dobrila University of Pula.

PROJECT OBJECTIVES:

The overall objective of the project is to enable the systematic testing of the perception and production of accents, taking into account the variability of speakers’ acquired accentual systems (both those originating from regions characterized by pitch-accent systems and those from stress-accent systems) and to develop a comprehensive audio repository of verbs. Building upon previous research and pedagogical experience in the research and teaching of the Croatian accentual system, the project aims to establish a robust methodological framework for the teaching of Croatian accentuation in a digital environment. The specific objectives are to advance open science through the dissemination of research results and the development of an open-access audio repository and related digital tools; to design and implement innovative digital solutions for the automatic recognition of Croatian accents (including both standard-language and stress-based varieties) and for the automatic annotation of accent placement; to promote science through applications that support the learning and teaching of the Croatian accentual norm, particularly in the areas of accent perception and production; to enhance academic programs in Croatian linguistics, phonetics, and related disciplines where mastery of standard speech is a professional requirement; to strengthen interdisciplinary research collaboration, especially between the fields of Croatian linguistics, phonetics, computer science, and information technology; and, finally, to establish partnerships with other research institutions for the implementation of large language models and speech transformers in the analysis and processing of speech, as well as in phonetic and sociophonetic research.

PROJECT HOLDER: University of Juraj Dobrila in Pula