Ulfeta Marovac
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Ulfeta Marovac has a total of 34 co-authors. Here are the top co-authors by number of shared publications:
- Aldina Avdić
Department of Technical and Technological Sciences, State University of Novi Pazar, Vuka Karadžića 9, Novi Pazar 36300, Serbia
10 shared publications - Hojjatollah Farahani
Tarbiat Modares University
6 shared publications - Kruna Ratković
University of Donja Gorica
6 shared publications - Marko Simeunović
University of Donja Gorica
5 shared publications - Nataša Kovač
University of Donja Gorica
5 shared publications - Dragan Jankovic
University of Nis
3 shared publications
Latest Publications
- 4 June 2026
- Published by Springer Nature
- p. 371-386
Bayesian Data Analysis
- 4 June 2026
- Published by Springer Nature
- p. 173-202
Data Modelling
- 4 June 2026
- Published by Springer Nature
- p. 387-407
Support Vector Machine Algorithm
- 4 June 2026
- Published by Springer Nature
- p. 447-472
Natural Language Processing and Psychological Sciences
- 4 June 2026
- Published by Springer Nature
- p. 447-472
Natural Language Processing and Psychological Sciences
- 4 June 2026
- Published by Springer Nature
- p. 153-171
The Difference Between Data Science, Statistics, Machine Learning, and Artificial Intelligence
- 4 June 2026
- Published by Springer Nature
- p. 153-171
The Difference Between Data Science, Statistics, Machine Learning, and Artificial Intelligence
- 4 June 2026
- Published by Springer Nature
- p. 409-445
Classification in Psychological Research
- 18 May 2026
- Published by University Library in Kragujevac in Scientific Publications Of The State University Of Novi Pazar, series A: Applied Mathematics, Informatics and Mechanics
- Vol. 26 (1)
Architecture and Evaluation of a Multi-Agent System for Market Monitoring
The contemporary financial environment is characterized by a rapidly increasing volume of heterogeneous data, which necessitates automated solutions for their integration and analysis in the context of market dynamics. In this paper, a multi-agent system for automated monitoring and analysis of capital markets is presented, developed using the CrewAI framework and combining structured financial data with sentiment analysis of financial news. The proposed system integrates numerical data with textual content from relevant news sources, enabling the identification of information with a potential impact on stock price movements. The system architecture comprises two specialized agents: MarketScanner, responsible for analyzing market indicators and generating concise analytical reports, and NewsAggregator, which collects relevant news and performs sentiment analysis using large language models and the FinBERT model for financial text processing. The system was evaluated using data from Apple, Bank of America, and Exxon Mobil. Experimental results indicate that the integration of a multi-agent architecture with sentiment analysis supports systematic examination of market-relevant information. Sentiment classification accuracy reached above 80% in selected cases when evaluated against expert annotations. The findings suggest that sentiment–market relationships vary across companies and depend on the specific market context.
- 18 May 2026
- Published by University Library in Kragujevac in Scientific Publications Of The State University Of Novi Pazar, series A: Applied Mathematics, Informatics and Mechanics
- Vol. 26 (1)
Architecture and Evaluation of a Multi-Agent System for Market Monitoring
The contemporary financial environment is characterized by a rapidly increasing volume of heterogeneous data, which necessitates automated solutions for their integration and analysis in the context of market dynamics. In this paper, a multi-agent system for automated monitoring and analysis of capital markets is presented, developed using the CrewAI framework and combining structured financial data with sentiment analysis of financial news. The proposed system integrates numerical data with textual content from relevant news sources, enabling the identification of information with a potential impact on stock price movements. The system architecture comprises two specialized agents: MarketScanner, responsible for analyzing market indicators and generating concise analytical reports, and NewsAggregator, which collects relevant news and performs sentiment analysis using large language models and the FinBERT model for financial text processing. The system was evaluated using data from Apple, Bank of America, and Exxon Mobil. Experimental results indicate that the integration of a multi-agent architecture with sentiment analysis supports systematic examination of market-relevant information. Sentiment classification accuracy reached above 80% in selected cases when evaluated against expert annotations. The findings suggest that sentiment–market relationships vary across companies and depend on the specific market context.
- 18 February 2026
- Published by Springer Nature
- p. 357-370
Generative Artificial Intelligence as a Tool for Improving the Accuracy of Classification Models
- 18 February 2026
- Published by Springer Nature
- p. 357-370
Generative Artificial Intelligence as a Tool for Improving the Accuracy of Classification Models
- 1 January 2024
- Published by University Library in Kragujevac in Scientific Publications Of The State University Of Novi Pazar, series A: Applied Mathematics, Informatics and Mechanics
- Vol. XVI (1)
Towards Citizen-Centered Smart Services Research Insights from a Project
This paper presents research on citizen-centered smart services from the perspective of a long-term project. Initially intended for E-Government services in social, economic and biological domains, it is naturally extended to smart city services due to the common goals of citizen-centered services, common data sources, similar technologies, and the like. It is also overlapped with some international cooperation with similar goals. Both, E-Government and Smart Cities, feed with information that originates from their sensors, citizens and IoT, and, in return, provide smart services that aim to improve the quality of life and achieve more efficient management in the different environments in which they operate. The sensing is provided by listening to citizen opinions via social media, like Facebook, X (formerly, Twitter), and the like, and other Web 2.0 methods like crowdsensing and crowdsourcing, etc. The emerging world of IoT allows to sense some environmental phenomena like temperature, air pollutants, traffic congestion, fire, flooding, and to evidence some illegal behavior on the streets, for example. Development and deployment of citizen-centered services are analyzed through the prism of four dimensions, namely (1) modeling and decision-making, (2) sensing and analyzing, (3) willingness and engagement, and (4) openness and transparency. In addition, key technological enablers that drive the full achievement of these dimensions are briefly discussed. To address the challenges several issues are discusses. These include the enormous amount of data with high velocity, value and variety, the complexity of Serbian language and grammar with the absence of linguistic resources, and the limitations of IoT resources, to mention a few. These challenges are discussed by related work and a brief overview of proposed solutions with special emphasis on a framework, design of several specialized social machines, text processing in Serbian language, sentiment analysis, business process management and an excerpt of developed and deployed services. We briefly outline the main contributions of published work, including the key and technologies utilized, all supported by several experimental results. We conclude by identifying some unresolved problems and proposing future research directions.
- 1 January 2024
- Published by University Library in Kragujevac in Scientific Publications Of The State University Of Novi Pazar, series A: Applied Mathematics, Informatics and Mechanics
- Vol. XVI (1)
Towards Citizen-Centered Smart Services Research Insights from a Project
This paper presents research on citizen-centered smart services from the perspective of a long-term project. Initially intended for E-Government services in social, economic and biological domains, it is naturally extended to smart city services due to the common goals of citizen-centered services, common data sources, similar technologies, and the like. It is also overlapped with some international cooperation with similar goals. Both, E-Government and Smart Cities, feed with information that originates from their sensors, citizens and IoT, and, in return, provide smart services that aim to improve the quality of life and achieve more efficient management in the different environments in which they operate. The sensing is provided by listening to citizen opinions via social media, like Facebook, X (formerly, Twitter), and the like, and other Web 2.0 methods like crowdsensing and crowdsourcing, etc. The emerging world of IoT allows to sense some environmental phenomena like temperature, air pollutants, traffic congestion, fire, flooding, and to evidence some illegal behavior on the streets, for example. Development and deployment of citizen-centered services are analyzed through the prism of four dimensions, namely (1) modeling and decision-making, (2) sensing and analyzing, (3) willingness and engagement, and (4) openness and transparency. In addition, key technological enablers that drive the full achievement of these dimensions are briefly discussed. To address the challenges several issues are discusses. These include the enormous amount of data with high velocity, value and variety, the complexity of Serbian language and grammar with the absence of linguistic resources, and the limitations of IoT resources, to mention a few. These challenges are discussed by related work and a brief overview of proposed solutions with special emphasis on a framework, design of several specialized social machines, text processing in Serbian language, sentiment analysis, business process management and an excerpt of developed and deployed services. We briefly outline the main contributions of published work, including the key and technologies utilized, all supported by several experimental results. We conclude by identifying some unresolved problems and proposing future research directions.
- Open Access
A Survey of Resources and Methods for Natural Language Processing of Serbian Language
The Serbian language is a Slavic language spoken by over 12 million speakers and well understood by over 15 million people. In the area of natural language processing, it can be considered a low-resourced language. Also, Serbian is considered a high-inflectional language. The combination of many word inflections and low availability of language resources makes natural language processing of Serbian challenging. Nevertheless, over the past three decades, there have been a number of initiatives to develop resources and methods for natural language processing of Serbian, ranging from developing a corpus of free text from books and the internet, annotated corpora for classification and named entity recognition tasks to various methods and models performing these tasks. In this paper, we review the initiatives, resources, methods, and their availability.
Open AccessA Survey of Resources and Methods for Natural Language Processing of Serbian Language
The Serbian language is a Slavic language spoken by over 12 million speakers and well understood by over 15 million people. In the area of natural language processing, it can be considered a low-resourced language. Also, Serbian is considered a high-inflectional language. The combination of many word inflections and low availability of language resources makes natural language processing of Serbian challenging. Nevertheless, over the past three decades, there have been a number of initiatives to develop resources and methods for natural language processing of Serbian, ranging from developing a corpus of free text from books and the internet, annotated corpora for classification and named entity recognition tasks to various methods and models performing these tasks. In this paper, we review the initiatives, resources, methods, and their availability.