Hybrid Event

Event program

Wednesday, 5/22/2024 4:00 PM - 7:00 PM, Camelia 1, Grand hotel Adriatic, Opatija 4:00 PM - 7:00 PM Papers  1. L. Brkić, I. Mekterović, M. Fertalj (University of Zagreb Faculty of Electrical Engineering and Computing, Zagreb, Croatia) Automated SQL Query Evaluations in Massive Database Courses  When assessing submissions in a massive course, using an Automated Programming Assessment System (APAS), can benefit both students and teachers. Students can expect fast and consistent assessment, while teachers benefit from a reduced workload. Acquiring proficiency in SQL is one of core goals of any introductory or advanced database course. Evaluation of students’ SQL queries differs from a general-purpose code evaluation, such as that for C or Java, by requiring a database on which the query will be evaluated and parameterized comparisons of the obtained recordsets. The evaluation using APAS is typically performed in such a way that the system executes two queries: the student query and the correct query provided by the course staff and compares the resulting datasets in terms of accuracy and completeness. When comparing the obtained datasets, there are some factors to consider, including the importance of tuple ordering and the relevance of column names. Moreover, some SQL statements create, alter, or delete database objects such as tables and indexes, and their correctness cannot be determined using a predefined output dataset. To do so, every SQL question in APAS must reference some test database (populated with data). With hundreds of students enrolled, it becomes technically challenging to execute and evaluate their queries in real-time, especially as these databases begin to pile up due to the development and evolution of the courses. In this paper, we comment on a possible solution and present our approach with APAS that uses multiple cloned instances of the test database while supporting the aforementioned specifics of SQL query evaluation. 2. E. Samsonova, Z. Morić (Sveučilište Algebra, Zagreb, Croatia), G. Gvozden, T. Hlupić (Poslovna inteligencija d. o. o., Zagreb, Croatia) Data Enrichment Pipeline Model for Web Site Classification  In the contemporary landscape marked by the digitization of sales and marketing, burgeoning online platforms, and the pervasive influence of e-commerce, the geographical constraints that once delimited businesses have dissipated, providing enterprises the capacity to operate on a global scale. Complex data pipelines and AI systems are nowadays utilized to facilitate advanced analytics, operational optimization, and enhanced decision-making. Despite the ubiquity of such technological advancements, smaller enterprises are restricted by constraints and limitations, e. g. financial constraints, human resource limitations, and infrastructural and knowledge deficiencies, hindering the adoption of technological solutions and efficient data utilization. In light of these challenges, this paper proposes a resource-efficient system utilizing an automated Business-to-Business (B2B) lead database enrichment pipeline featuring a machine-learning solution for classification. The proposed machine-learning model proffers a low-complexity framework, mitigating resource and computational demands. The lead enrichment pipeline proposed in this paper encompasses components such as Customer Relationship Management (CRM) software, a web data extraction tool, a machine learning model for website classification, cloud-based database infrastructure, and a scheduling and orchestration tool, putting an emphasis on open-source and cost-effective solutions. The discussion concludes with a brief discussion of GDPR and privacy regulations. 3. A. Terović, I. Mekterović (University of Zagreb, Faculty of Electrical Engineering and Computing, Zagreb, Croatia) On Improving the Qualitative Features of the User Interface of Mobile Applications using Machine Learning Methods  User interfaces are among the most frequently used systems for interaction. To ease the process of creating user interfaces for designers and developers, in this paper, we aim to explore methods for improving interface design. We propose a methodology focused on improving UI through the implementation of a layout generation model. This model leverages Diffusion Layout Transformer (DLT) and is trained using comprehensive datasets such as Rico, Clay and a Huggingface dataset. The effectiveness of our model is evaluated based on its ability to generate aesthetically pleasing and functional UI layouts. We conclude the paper by discussing the implications of our findings and outlining future research directions in the automation of UI design using machine learning. The paper underscores the potential and challenges of integrating machine learning in UI design, paving the way for future advancements in automated UI layout generation. 4. A. Babić, A. Zron (Faculty of Economics and Business, University of Rijeka, Rijeka, Croatia) Business Intelligence Tools in the Interpretation of the Ranking of Smart Cities  In the age of big data, it is necessary to use various tools to quickly identify data trends, i.e. to present large data sets in a graphical format. This paper explains how data is extracted, transformed and loaded (ETL) from different sources, i.e. it describes how business intelligence tools are used in interpreting a large data set for a large number of cities. Three basic phases of collection, processing and visualization of 38 indicators by six dimensions of a smart city for 127 Croatian cities are presented. Data collection was done through web scraping, secondary searches on the official websites of public institutions and through official requests to authorized institutions and cities. The second phase involved structuring, normalization, formatting, data cleaning and calculation of the index of smart Croatian cities in ETL, which is formed as an average of all indicator values for all six dimensions of a smart city. The third phase concerned the interpretation of the results, i.e. the visualization for all dimensions of smart cities in Power BI and Tableau. 5. I. Mekterović, L. Brkić, M. Fertalj (University of Zagreb, Faculty of Electrical Engineering and Computing, Zagreb, Croatia , Zagreb, Croatia), M. Fabijanić (Algebra University College/Software Engineering, Zagreb, Croatia, Zagreb, Croatia) Interactive Programming Tutorials in Automated Programming Assessment System Edgar  Automated programming assessment tools are software systems widely used in education to assess programming code without manual intervention. Beyond exam scenarios, these tools are increasingly applied in e-learning contexts. In this realm, interactive programming tutorials have gained prominence for their effectiveness in teaching programming concepts. These tutorials blend theoretical knowledge with hands-on exercises, providing real-time feedback on code errors to facilitate prompt identification and correction by learners. The interactive nature engages learners actively, enhancing their understanding, and adaptability accommodates individualized progress. The accessibility and scalability of interactive programming tutorials suit learners of diverse skill levels. Integrated with automated assessment systems, these tutorials not only provide a dynamic and personalized learning experience but also alleviate the burden on instructors by enabling interactive content creation and offering valuable learning analytics. This paper introduces an evolution of the Edgar system, now equipped with an integrated interactive tutorial module. This module can evaluate embedded questions and code playgrounds in various programming languages, including SQL, Java, C, Python, etc., as well as multiple-choice questions. The integration represents a practical shift in programming education, offering learners a versatile and personalized approach to acquiring essential skills. 6. D. Mlinarić, J. Dončević, M. Brčić, I. Botički ( Faculty of Electrical Engineering and Computing, Zagreb, Croatia) Revolutionizing Software Development: Autonomous Software Evolution  This paper discusses a software development method focused on creating a self-adapting and evolving system using AI and ML techniques. The goal is to reduce the need for manual software updates, offering a solution that continuously adapts to changing requirements. To achieve this goal, an AI-based update model is presented, and a possible system is discussed. Use case example demonstrate the applicability of the update model in real-world scenarios. As a cloud-based solution, this method could ensure scalability and broad applicability across various industry sectors. 7. N. Caporusso, A. Campan, A. Khanal, G. Upreti, M. Doan (Northern Kentucky University, Highland Heights, United States) Competitive Arousal, Temporal Clustering, and Overbidding Dynamics in an Auction Platform  Auctions provide a unique lens through which human behavior can be studied, particularly in decision-making, strategy, and economic interaction. The data generated in auction environments, characterized by competitive bidding, time constraints, and varying levels of information asymmetry, offers rich insights into how individuals make decisions under pressure, uncertainty, and scarcity. Researchers can gain a deeper understanding of cognitive biases and strategic decision-making in competitive environments by analyzing bid increments, winning bids, and participant behavior across various auction types. In this paper, we present the results of a preliminary study that focused on a dataset from a popular online auction website where users can bid on different types of retail products. Our findings report three interconnected and apparently contradictory auction dynamics discovered in the dataset, including irrational behavior, highlighting the significance of our research and the need for additional future studies. 8. M. Morgen, L. Begic Fazlic, G. Dartmann (Trier University of Applied Sciences - Environmental Campus Birkenfeld, Hoppstädten-Weiersbach, Germany) Connect, Understand and Learn: Dynamic Knowledge Graph Transforms Learning  The automation of knowledge bases is a challenge if only small training data sets are available for the corresponding learning methods. The approach presented in this paper can work with small training data sets and enables the solution of tasks with previously hidden syntactic structures. In this research, a new conceptual algorithm for learning and updating knowledge bases is proposed. We combined a powerful NLP approach with statistical methods to build a word frequency-based corpus for various question answering problems. Then, we used specific similarity measures to find the best solution for the given problem. For this purpose, a vector model is used and weights are calculated for the association between terms and problems. In the last phase, we created a continuous learning model with a dynamic knowledge graph that can be updated with a new task and predict an answer to upcoming problems. The knowledge base is updated with new information when the pattern of a problem is unknown and therefore not found. The implementation of the algorithm is validated using various openly available data sets from the field of user support in business and medicine. The proposed method supports an incremental learning approach and real-time implementation. 9. A. Tanovic, A. Cerimagic-Hasibovic (Faculty of Electrical Engineering, University of Sarajevo, Sarajevo, Bosnia and Herzegovina) Automated Real Estate Chatbot  Fully automated chatbots are increasingly being applied in the real-estate industry. Although, they are not completely able replace interaction between the real-estate agents, they can automate customer support, save human resources for qualitative tasks, accelerate operations, and improve business branding. In this paper, a chatbot for real-estate is developed. The chatbot is able to engage clients in meaningful conversations in real-time. It provides a 24/7 service and effectively reduces administrative costs. The architecture and infrastructure overview are presented. The rule matching algorithm is presented and discussed in detail. 10. R. Pavliš (University of Zagreb, Faculty of Electrical Engineering and Computing, Zagreb, Croatia) Graph Databases: An Alternative to Relational Databases in an Interconnected Big Data Environment  As the global volume of data continues to rise at an unprecedented rate, the challenges of storing and analyzing data are becoming more and more highlighted. This is especially apparent when the data are heavily interconnected. The traditional methods of storing and analyzing data such as relational databases often encounter difficulties when dealing with large amounts of data and this is even more pronounced when the data exhibits intricate interconnections. This paper introduces graph databases as an alternative to relational databases in an interconnected Big Data environment. It will also show the theoretical basis behind graph databases and how they outperform relational databases in such an environment, but also why they are better suited for this kind of environment than other NoSQL alternatives. A state of the art in graph databases and how they compare to relational databases in various scenarios will also be presented in this paper. 11. A. Stojanović, S. Plehati, Ž. Kovačević (Tehničko veleučilište u Zagrebu, Zagreb, Croatia) A Domain-Specific Language Implementation Framework for C++ Based on S-expressions  Domain-specific languages (DSLs) are languages designed and implemented for a specific application domain. Designing and implementing such languages is hard because it requires both domain-specific and language implementation knowledge. However, they offer its users substantial gains in productivity because they consist of constructs that directly represent domain-specific concepts. The absence of extensible and easy-to-use domain-specific language frameworks for C++ has made it difficult for developers to build such languages. Designing syntax, building parsers and then implementing the execution environment is time consuming, error prone and requires lots of work. In this paper we present a C++ framework for implementing domain-specific languages with predefined syntax based on S-expressions, but avoiding the strict parenthesization required by them. The framework offers two advantages over implementing a domain-specific language “by hand”: 1) the syntax is already built into the framework so users do not need to design and implement it, and 2) the framework provides the execution environment where users just need to implement domain-specific commands by extending the command set of the interpreter. This framework can also be used for building command-line interpreters for automation, control and/or configuration of software systems, or as a textual interface for software tools and applications. 12. V. Bosilj Vukšić, L. Ivančić, D. Suša Vugec (Faculty of Economics and Business, University of Zagreb, Zagreb, Croatia) Process Mining: Application in Practice  Process mining is a concept that is not new, and software in this area is well represented on the market. The academic interest is reflected in a relatively large number of articles dealing with this topic. However, the practical implementation does not quite match these development trends. The aim of this study is therefore to investigate (1) the industries in which process mining is used and (2) the specific software used for this purpose. For the study, a systematic literature review was conducted by examining articles published in the last 5 years in Clarivate Analytics’ Web of Science (WoS) database. Based on the collected results, a bibliographic analysis was performed, the case studies on the application areas of process mining were analyzed and the software used was systematized. In addition, the identified benefits, current status, trends and challenges in the implementation of this concept in practice were analyzed. Considering the limitations of the research, guidelines for the scientific community are proposed to guide future research. Furthermore, approaches are proposed that could contribute to a stronger application of process mining in business practice.

Basic information:

Boris Vrdoljak (Croatia), Matteo Golfarelli (Italy), Mihaela Vranić (Croatia)

Marko Banek (Croatia), Matteo Golfarelli (Italy), Damir Pintar (Croatia), Mihaela Vranić (Croatia), Boris Vrdoljak (Croatia)

Mirta Baranović (Croatia), Ladjel Bellatreche (France), Ljiljana Brkić (Croatia), Alfredo Cuzzocrea (Italy), Matteo Francia (Italy), Enrico Gallinucci (Italy), Paolo Garza (Italy), Marko Gulić (Croatia), Luka Humski (Croatia), Igor Mekterović (Croatia), Sandro Skansi (Croatia), A Min Tjoa (Austria)

Registration / Fees:

The discount doesn't apply to PhD students.

NOTE FOR AUTHORS: In order to have your paper published, it is required that you pay at least one registration fee for each paper. Authors of 2 or more papers are entitled to a 10% discount.

Boris Vrdoljak
University of Zagreb
Faculty of Electrical Engineering and Computing
Unska 3
HR-10000 Zagreb, Croatia

Phone: +385 1 6129 532
Fax: +385 1 6129 915
E-mail: boris.vrdoljak@fer.hr

The best papers will get a special award.
Accepted papers will be published in the ISSN registered conference proceedings. Papers in English presented at the conference will be submitted for inclusion in the IEEE Xplore Digital Library.
.............
There is a possibility that the selected scientific papers with some further modification and refinement are being published in the following journals: Journal of Computing and Information Technology (CIT), MDPI Applied Science, MDPI Information Journal, Frontiers and EAI Endorsed Transaction on Scalable Information Systems.