Lucia Nacinovic Prskalo
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Lucia Nacinovic Prskalo has a total of 11 co-authors. Here are the top co-authors by number of shared publications:
- Marija Brkic Bakaric
Faculty of Informatics and Digital Technologies, University of Rijeka, 51000 Rijeka, Croatia
9 shared publications Iva Pauletic
1 shared publication- Katarina Brkljača
Erste&Steiermärkische Bank d.d.
1 shared publication Košuta Estera Lerga
1 shared publicationKristina Tonkovic
Department of Informatics University of Rijeka Rijeka Croatia
1 shared publicationM. Brkic Bakaric
University of Rijeka,Faculty of Informatics and Digital Technologies,Rijeka,Croatia
1 shared publication
Latest Publications
- 1 January 2025
- Published by University of Zadar
Adapting the Generic English-Croatian NMT Model to a Religious Domain
Recent discoveries in the field of artificial intelligence have significantly impacted various professions, including the translation industry, leading to notable changes in translators’ work processes. The study presented in this article indicates that today any translator, even those without advanced IT skills, can develop a higher quality Neural Machine Translation (NMT) system based on their own texts. This paper evaluates Google’s AutoML Translation service, which enables users to train high-quality models using their own text data. Specifically, AutoML Translation integrates an additional layer that tailors the generic Translation API model to a specific domain. The training process involves providing a user-defined dataset containing aligned sentences in the source and target languages. Google’s AutoML Translation service was used to adapt the base English-Croatian Google NMT model to the field of religion. Following a brief introduction to machine translation, this paper outlines the key aspects of the training and evaluation processes. Additionally, it presents two corpora employed in the training phase. The results demonstrate that a customized model outperforms the base model, as evidenced by the BLEU score.
- 1 January 2025
- Published by University of Zadar
Adapting the Generic English-Croatian NMT Model to a Religious Domain
Recent discoveries in the field of artificial intelligence have significantly impacted various professions, including the translation industry, leading to notable changes in translators’ work processes. The study presented in this article indicates that today any translator, even those without advanced IT skills, can develop a higher quality Neural Machine Translation (NMT) system based on their own texts. This paper evaluates Google’s AutoML Translation service, which enables users to train high-quality models using their own text data. Specifically, AutoML Translation integrates an additional layer that tailors the generic Translation API model to a specific domain. The training process involves providing a user-defined dataset containing aligned sentences in the source and target languages. Google’s AutoML Translation service was used to adapt the base English-Croatian Google NMT model to the field of religion. Following a brief introduction to machine translation, this paper outlines the key aspects of the training and evaluation processes. Additionally, it presents two corpora employed in the training phase. The results demonstrate that a customized model outperforms the base model, as evidenced by the BLEU score.
- Open Access
- 13 December 2024
- Published by Association for Computing Machinery (ACM)
- p. 136-141
Fine-tuning BERT for the task of metaphorical collocations identification
- 13 December 2024
- Published by Association for Computing Machinery (ACM)
- p. 136-141
Open AccessFine-tuning BERT for the task of metaphorical collocations identification
- 1 March 2024
- Published by IATED Academy in INTED Proceedings
- p. 3752-3760
EMPOWERING EDUCATION: DESIGN AND IMPLEMENTATION OF A WEB APP FOR MONITORING STUDENTS' PROFESSIONAL PRACTICE
- 1 March 2024
- Published by IATED Academy in INTED Proceedings
- p. 3752-3760
EMPOWERING EDUCATION: DESIGN AND IMPLEMENTATION OF A WEB APP FOR MONITORING STUDENTS' PROFESSIONAL PRACTICE
- 1 March 2024
- Published by IATED Academy in INTED Proceedings
- Vol. 1, 3836-3843
BRIDGING THE GAP: ENHANCING PARENT-EDUCATOR COMMUNICATION IN PRESCHOOLS WITH “KINDERGARTEN CONNECT” APPLICATION
- 1 March 2024
- Published by IATED Academy in INTED Proceedings
- Vol. 1, 3836-3843
BRIDGING THE GAP: ENHANCING PARENT-EDUCATOR COMMUNICATION IN PRESCHOOLS WITH “KINDERGARTEN CONNECT” APPLICATION
- Open Access
- 1 January 2024
- Published by Polytechnic of Rijeka University in Zbornik Veleučilišta u Rijeci
- Vol. 12 (1) , 375-393
Pregled prednosti i nedostataka digitalizacije, digitalnih rješenja, tehnologija i aplikacija korištenih u vrijeme pandemije
Ljudsko djelovanje u 21. stoljeću, pod utjecajem je 4. industrijske revolucije koja na dnevnoj bazi mijenja način na koji društvo živi, napreduje i egzistira. Neočekivanom pojavom pandemije, kao pozitivan odgovor, uslijedila su brojna digitalna rješenja koja su potaknula još veći razvoj i promet, kako web tako i mobilnih aplikacija te ostalih digitalnih rješenja. Glavni cilj ovog rada je istražiti postupke digitalizacije, digitalnih rješenja, tehnologija i aplikacija koje su se koristile tijekom pandemije te ukazati na napredak digitalnih rješenja razvijenih u novim, nepoznatim okolnostima kao i identificirati moguće izazove u njihovom daljnjem razvoju i primjeni te predložiti moguća rješenja. U radu se opisuje uloga tehnologije u olakšavanju svakodnevnih aktivnosti u izvanrednim okolnostima te se identificiraju prednosti i mogući izazovi u primjeni digitalnih rješenja u takvim okolnostima. U tu se svrhu opisuju najutjecajnije web aplikacije i digitalne tehnologije te njihova uloga u osiguranju normalnog tijeka života u izazovnim vremenima.
- 1 January 2024
- Published by Polytechnic of Rijeka University in Zbornik Veleučilišta u Rijeci
- Vol. 12 (1) , 375-393
Open AccessPregled prednosti i nedostataka digitalizacije, digitalnih rješenja, tehnologija i aplikacija korištenih u vrijeme pandemije
Ljudsko djelovanje u 21. stoljeću, pod utjecajem je 4. industrijske revolucije koja na dnevnoj bazi mijenja način na koji društvo živi, napreduje i egzistira. Neočekivanom pojavom pandemije, kao pozitivan odgovor, uslijedila su brojna digitalna rješenja koja su potaknula još veći razvoj i promet, kako web tako i mobilnih aplikacija te ostalih digitalnih rješenja. Glavni cilj ovog rada je istražiti postupke digitalizacije, digitalnih rješenja, tehnologija i aplikacija koje su se koristile tijekom pandemije te ukazati na napredak digitalnih rješenja razvijenih u novim, nepoznatim okolnostima kao i identificirati moguće izazove u njihovom daljnjem razvoju i primjeni te predložiti moguća rješenja. U radu se opisuje uloga tehnologije u olakšavanju svakodnevnih aktivnosti u izvanrednim okolnostima te se identificiraju prednosti i mogući izazovi u primjeni digitalnih rješenja u takvim okolnostima. U tu se svrhu opisuju najutjecajnije web aplikacije i digitalne tehnologije te njihova uloga u osiguranju normalnog tijeka života u izazovnim vremenima.
- 1 January 2023
- Published by Institute of Croatian Language and Linguistics in Rasprave Instituta za hrvatski jezik i jezikoslovlje
- Vol. 49 (1) , 1-32
Insights Into Automatic Extraction of Metaphorical Collocations
Collocations have been the subject of much scientific research over the years. The focus of this research is on a subset of collocations, namely metaphorical collocations. In metaphorical collocations, a semantic shift has taken place in one of the c...
- 1 January 2023
- Published by Institute of Croatian Language and Linguistics in Rasprave Instituta za hrvatski jezik i jezikoslovlje
- Vol. 49 (1) , 1-32
Insights Into Automatic Extraction of Metaphorical Collocations
Collocations have been the subject of much scientific research over the years. The focus of this research is on a subset of collocations, namely metaphorical collocations. In metaphorical collocations, a semantic shift has taken place in one of the c...
- Open Access
- 10 November 2022
- Published by MDPI AG in Electronics
- Vol. 11 (22) , 3687
System for Automatic Assignment of Lexical Stress in Croatian
It is very popular today to integrate voice interfaces into IoT devices. The pronunciation and proper prosody of speech play a major role in the intelligibility and naturalness of synthesized voices. Each language has its own prosodic characteristics. In this paper, we present the results of a study aimed at testing the applicability of methods for modelling and predicting the prosodic features of the Croatian language. The extent to which their performance can be improved by incorporating linguistic features and linguistic peculiarities specific to the Croatian language was investigated. In the model learning process, tree classification was used to predict the lexical stress position and the type of stress in a word, and a lexicon of 1,011,785 word forms was used as the model learning set. Separate models were created for predicting the position and type of lexical stress. The results improved significantly after the rules for atonic words (clitics) were applied. A hybrid approach combining a rule-based approach and a modelling approach was also proposed. The final accuracy of assigning lexical stress using the hybrid approach was 95.3%.
- 10 November 2022
- Published by MDPI AG in Electronics
- Vol. 11 (22) , 3687
Open AccessSystem for Automatic Assignment of Lexical Stress in Croatian
It is very popular today to integrate voice interfaces into IoT devices. The pronunciation and proper prosody of speech play a major role in the intelligibility and naturalness of synthesized voices. Each language has its own prosodic characteristics. In this paper, we present the results of a study aimed at testing the applicability of methods for modelling and predicting the prosodic features of the Croatian language. The extent to which their performance can be improved by incorporating linguistic features and linguistic peculiarities specific to the Croatian language was investigated. In the model learning process, tree classification was used to predict the lexical stress position and the type of stress in a word, and a lexicon of 1,011,785 word forms was used as the model learning set. Separate models were created for predicting the position and type of lexical stress. The results improved significantly after the rules for atonic words (clitics) were applied. A hybrid approach combining a rule-based approach and a modelling approach was also proposed. The final accuracy of assigning lexical stress using the hybrid approach was 95.3%.
- 16 September 2022
- Published by Springer Nature
- p. 102-112
Identification of Metaphorical Collocations in Different Languages – Similarities and Differences
- 16 September 2022
- Published by Springer Nature
- p. 102-112
Identification of Metaphorical Collocations in Different Languages – Similarities and Differences
- 23 May 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 568-572
The LanGuide Project - Strengths and Threats
The project "Language guidance tool for improving language knowledge" (LanGuide) is supported by the Erasmus+ program Strategic Partnerships for Higher Education. The project duration is 36 months. The coordinating institution is the University of Primorska, and the partner institutions are Mälardalen University, the University of Castilla - La Mancha, Transilvania University of Brașov, and two Croatian universities — the University of Zadar, and the University of Rijeka. The main aim of the project is to build an open access students/teachers/administrative staff centered mobile app for improving language knowledge. To achieve the goal, the first phase of the project is concerned with the development of a universal methodology for the preparation of language exercises focused on language for specific purposes. In the second phase, the methodology is applied in the preparation of language resources for six languages in four LSP fields at different levels of difficulty. Appropriate IT tools are developed simultaneously with the preparation of exercises. Finally, the developed tools and resources are evaluated through two rounds of user testing. This paper first describes the motivation behind the project proposal. Next, each of the four intellectual outputs and activities carried out within the project are described. Finally, challenges such as the one brought about by the pandemic of COVID-19 are highlighted.
- 23 May 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 568-572
The LanGuide Project - Strengths and Threats
The project "Language guidance tool for improving language knowledge" (LanGuide) is supported by the Erasmus+ program Strategic Partnerships for Higher Education. The project duration is 36 months. The coordinating institution is the University of Primorska, and the partner institutions are Mälardalen University, the University of Castilla - La Mancha, Transilvania University of Brașov, and two Croatian universities — the University of Zadar, and the University of Rijeka. The main aim of the project is to build an open access students/teachers/administrative staff centered mobile app for improving language knowledge. To achieve the goal, the first phase of the project is concerned with the development of a universal methodology for the preparation of language exercises focused on language for specific purposes. In the second phase, the methodology is applied in the preparation of language resources for six languages in four LSP fields at different levels of difficulty. Appropriate IT tools are developed simultaneously with the preparation of exercises. Finally, the developed tools and resources are evaluated through two rounds of user testing. This paper first describes the motivation behind the project proposal. Next, each of the four intellectual outputs and activities carried out within the project are described. Finally, challenges such as the one brought about by the pandemic of COVID-19 are highlighted.
- Open Access
- 1 January 2020
- Published by The Science and Information Organization in International Journal of Advanced Computer Science and Applications
- Vol. 11 (5)
Clash between Segment-level MT Error Analysis and Selected Lexical Similarity Metrics
The aim of this paper is to evaluate the quality of popular machine translation engines on three texts of different genre in a scenario in which both source and target languages are morphologically rich. Translations are obtained from Google Translate and Microsoft Bing engines and German-Croatian is selected as the language pair. The analysis entails both human and automatic evaluation. The process of error analysis, which is time-consuming and often tiresome, is conducted in the user-friendly Windows 10 application TREAT. Prior to annotation, training is conducted in order to familiarize the annotator with MQM, which is used in the annotation task, and the interface of TREAT. The annotation guidelines elaborated with examples are provided. The evaluation is also conducted with automatic metrics BLEU and CHRF++ in order to assess their segment-level correlation with human annotations on three different levels–accuracy, mistranslation, and the total number of errors. Our findings indicate that neither the total number of errors, nor the most prominent error category and subcategory, show consistent and statistically significant segment-level correlation with the selected automatic metrics.
- 1 January 2020
- Published by The Science and Information Organization in International Journal of Advanced Computer Science and Applications
- Vol. 11 (5)
Open AccessClash between Segment-level MT Error Analysis and Selected Lexical Similarity Metrics
The aim of this paper is to evaluate the quality of popular machine translation engines on three texts of different genre in a scenario in which both source and target languages are morphologically rich. Translations are obtained from Google Translate and Microsoft Bing engines and German-Croatian is selected as the language pair. The analysis entails both human and automatic evaluation. The process of error analysis, which is time-consuming and often tiresome, is conducted in the user-friendly Windows 10 application TREAT. Prior to annotation, training is conducted in order to familiarize the annotator with MQM, which is used in the annotation task, and the interface of TREAT. The annotation guidelines elaborated with examples are provided. The evaluation is also conducted with automatic metrics BLEU and CHRF++ in order to assess their segment-level correlation with human annotations on three different levels–accuracy, mistranslation, and the total number of errors. Our findings indicate that neither the total number of errors, nor the most prominent error category and subcategory, show consistent and statistically significant segment-level correlation with the selected automatic metrics.