Monday Computer Science Seminar Archives - UP Famnit

Monday Computer Science Seminar (MON, 19.10.2026 at 4 PM in VP3)

branko.kavsek — Mon, 19 Oct 2026 07:28:23 +0000

PRESENTER: Kaja PRAPROTNIK, a PhD student in Computer Science at UP FAMNIT. TITLE:Influencing Sentiment in LLM Output via Token Selection Manipulation (part 2) ABSTRACT:We conduct a human evaluation of sentiment steering in Llama-3.1-8B-Instruct, using arena-style pairwise comparison in which evaluators choose which of two outputs from the same prompt feels more positive or negative. The steering biases the logits of tokens drawn from a sentiment lexicon at each decoding step, without additional training, comparing an additive bias with a multiplication and division formulation. Outputs are also scored automatically with the lexicon-based sentiment analysis tool VADER, and analysed with Friedman tests and Wilcoxon signed-rank comparisons against the baseline. ABOUT THE PRESENTER:Kaja Praprotnik is a PhD student at UP FAMNIT under the supervision of Prof. Jernej Vičič. Her research focuses on natural language processing with a strong focus on artificial intelligence. PRESENTER: Jovan PAVLOVIĆ, PhD student in Computer Science at UP FAMNIT. TITLE:A retrieval-augmented generation framework designed to improve evidence retrieval for complex, multi-hop reasoning tasks ABSTRACT:This presentation focuses on a retrieval-augmented generation (RAG) framework designed to improve evidence retrieval for complex, multi-hop reasoning tasks. Conventional RAG systems often treat all retrieved passages as equally reliable. GraphRAG approaches augment retrieval with knowledge graphs that represent relationships between entities and support multi-step reasoning, but they may rely on pre-existing, external knowledge graphs and LLM decision-making during the retrieval process. Our method, called SA-RAG, aims to improve knowledge retrieval using a spreading activation algorithm to search for relevant information in an interconnected document corpus, without relying on an LLM to guide the process or on human-crafted knowledge graphs. Experiments on multi-hop question-answering tasks show that SA-RAG performs comparably to or better than state-of-the-art methods and improves answer correctness by approximately 12 percentage points over naive vector-based retrieval. ABOUT THE PRESENTER:Jovan Pavlović is a Ph.D. student and teaching assistant at UP FAMNIT, with research interests spanning network science and machine learning, graph-enhanced retrieval-augmented generation, and graph neural networks. Both seminars will be held in English, starting at 4 PM in the classroom FAMNIT-VP3.

The post Monday Computer Science Seminar (MON, 19.10.2026 at 4 PM in VP3) appeared first on UP Famnit.

“Eno morje idej 2026” Entrepreneurship Competition (MON, October 12, 2026, at 4:00 PM)

branko.kavsek — Fri, 09 Oct 2026 09:48:14 +0000

The Municipality of Izola, in cooperation with the Public Institute for the Promotion of Entrepreneurship and Development Projects of the Municipality of Izola, is once again organizing the competition for the most promising business ideas, “ENO MORJE IDEJ 2026”. The competition aims to promote innovation and entrepreneurship in the Municipality of Izola and beyond by selecting the most promising business ideas that will significantly contribute to the development of new projects, the establishment of companies, and networking. During the lecture, students will be introduced to the competition, the application procedure, and the conditions for participation. ABOUT THE PRESENTER:Dr. Tina B. Kregar leads the project to establish the Morje Business Incubator in Izola within the Public Institute for the Promotion of Entrepreneurship and Development Projects of the Municipality of Izola (JZP Izola). Her work focuses on strengthening entrepreneurial competencies, promoting innovation, accelerating business growth, and introducing sustainable business practices into the local community. During the seminar, she will present the entrepreneurship competition “Eno morje idej 2026”, organized by the Municipality of Izola in cooperation with JZP Izola to promote entrepreneurship and innovation in Izola. The lecture will take place as part of the Monday Computer Science Seminar on Monday, October 12, 2026, starting at 4:00 PM in lecture room FAMNIT-VP3 (in Slovene). More about the competition (Slovenian pamphlet).

The post “Eno morje idej 2026” Entrepreneurship Competition (MON, October 12, 2026, at 4:00 PM) appeared first on UP Famnit.

AI-Generated Text Detection for Under-Resourced Languages (Inaugural Lecture)

branko.kavsek — Sat, 03 Oct 2026 08:18:15 +0000

The rapid development of large language models (LLMs) has created a growing need for reliable methods to distinguish AI-generated text from human-written text. Existing detection approaches include statistical methods, stylometric analysis, perplexity-based techniques, transformer-based classifiers, and large neural models trained specifically for AI-text detection. Although many of these methods achieve high accuracy on benchmark datasets, their performance often decreases when applied to new domains, new language models, paraphrased texts, or modified content. Another important limitation is language bias. Most existing detectors are developed and evaluated mainly for English, while their performance is often much lower for other languages, especially for morphologically rich and low-resource languages. This talk investigates a simple question: how do existing detectors deal with under-represented (and in the era of LLMs all non-English languages belong to this group) and morphologically rich languages. By focusing on simple statistical properties of embedding representations, the proposed method offers a more language-independent alternative for multilingual AI-text detection, particularly for languages that are underrepresented in current research. Presenter: prof. dr. Jernej Vičič Jernej is a professor at the Faculty of Mathematics, Natural Sciences and Information Technologies of the University of Primorska. He completed his studies in Computer Science, where he developed research interests in language technologies, machine learning, distributed systems, and network analysis. His current work focuses on artificial intelligence, natural language processing, distributed computing, and blockchain-based systems. He collaborates on several national and international research projects related to AI, data analysis, and advanced digital technologies. The presentation will take place as part of the Monday Computer Science Seminaron Monday, 5.10.2026, starting at 16:00 in the lecture room FAMNIT-VP3.

The post AI-Generated Text Detection for Under-Resourced Languages (Inaugural Lecture) appeared first on UP Famnit.

Nadun LIYANAGE: Passenger Perception of Motion Under Visual Stimuli

vanja.kovse — Mon, 14 Sep 2026 10:37:45 +0000

On Monday, 14 September 2026, at 4:00 p.m., a lecture will be held as part of the MONDAY RIN SEMINAR, organized by the Departments of Information Sciences and Technologies at UP FAMNIT and UP IAM. TIME/VENUE: 14 September 2026 at 4:00 p.m., FAMNIT-VP3. Lecturer: Nadun LIYANAGE——————————————– Nadun Liyanage is a doctoral researcher at University of Primorska – HICUP Lab. His work explores how immersive technologies can improve interactions in vehicles and transportation contexts, such as driver assistance, infotainment, and user experience in future mobility systems. ————————————————————————————–Title: Passenger Perception of Motion Under Visual Stimuli————————————————————————————– Abstract: This study investigates how different visual stimuli influence passengers’ perception of vehicle motion during real-world travel. Participants experienced controlled visual conditions through a head-mounted display while their perceived lateral motion was continuously recorded and compared with actual vehicle movement. The study explores how visual information alters motion perception, with implications for designing comfortable and immersive passenger experiences in future autonomous vehicles. Seminar bo potekal v angleškem jeziku s pričetkom ob 16:00 v FAMNIT-VP3.

The post Nadun LIYANAGE: Passenger Perception of Motion Under Visual Stimuli appeared first on UP Famnit.

Jani SUBAN and Nilukshan KRISHNARAM

mina.djordjevic — Mon, 07 Sep 2026 07:52:17 +0000

On Monday, 7 September 2026, at 16:00, two lectures will be held as part of the MONDAY SEMINAR OF COMPUTER SCIENCE AND INFORMATICS organized by the Departments of Information Sciences and Technologies of UP FAMNIT and UP IAM. TIME/LOCATION: 7 Septemebr 2026 at 16:00 via Zoom. 1st lecture============ ————————————LECTURER: Jani SUBAN———————————— Jani Suban is a first-year PhD student and teaching assistant at UP FAMNIT and UL FRI. His research field is algorithms and data structures, with a focus on algorithms on strings. —————————————————————————————————-TITLE: Knight-merge: string concatenation becomes suffix tree merging—————————————————————————————————- ABSTRACT: Suffix trees are a widely used data structure for text indexing and pattern searching. A suffix tree is a path-compressed trie constructed from all suffixes of the input text. Its construction can be performed in linear time with respect to the length of the input text. This raises a new question: can we merge two suffix trees in linear time?We present a new algorithm, called Knight-merge, for merging two suffix trees into a single suffix tree. The algorithm runs in linear time, $O(n_1 + n_2)$, where $n_1$ and $n_2$ are the lengths of the two strings from which the suffix trees were constructed. It reuses the existing suffix trees and adds new nodes only when necessary. The algorithm preserves and computes suffix links, and it improves the time complexity of merging two suffix trees from $O((n_1 + n_2)^2)$ to $O(n_1 + n_2)$. The seminar will be held in English. ============================================================================================================= 2nd lecture:============ —————————————————–LECTURER: Nilukshan KRISHNARAM—————————————————– Nilukshan Krishnaram is a PhD student and teaching assistant at UP FAMNIT, where he is affiliated with the HICUP Lab. He holds a B.Sc. (Hons) in Information Systems from the University of Colombo, Sri Lanka. His research interests focus on human–computer interaction, extended reality, user experience design, and visual and perceptual attention guidance. ————————————————————————————————————————————TITLE: Look Here! But Did It Help? What Eye Tracking Reveals About Visual Attention Guidance———————————————————————————————————————————— ABSTRACT: Visual attention-guidance techniques are designed to help users find what matters. But, does a cue that grabs attention actually help users look in the right place? To investigate this question, we conducted an exploratory user study comparing paper-informed replications of 18 existing attention-guidance techniques within the same controlled visual-search task. Twelve participants completed 76 adaptive pairwise comparisons of these techniques while we collected subjective judgments alongside eye-tracking data, examining noticeability, preference, perceived findability, target-fixation success, guidance-relative time to first fixation, and replay effort. The results reveal that these measures tell very different stories: highly noticeable techniques were not necessarily the fastest or most reliable at guiding gaze, the most preferred techniques were not always the strongest performers, and some less-preferred techniques still achieved high search success. The findings show that no single measure can fully capture how well an attention-guidance technique works, highlighting the importance of considering both subjective judgments and gaze-based performance. The seminar will be held in English. ============================================================================================================= The seminars will be held online via the Zoom application, starting at 16:00, at the following link: https://upr-si.zoom.us/j/297328207?pwd=S3Zpdk1VR3pjckNtWkQwKzlvcDR5UT09 Meeting ID: 297 328 207Passcode: 123456789 You are cordially invited!

The post Jani SUBAN and Nilukshan KRISHNARAM appeared first on UP Famnit.

Matic POŽAR and Uroš SERGAŠ

vanja.kovse — Mon, 31 Aug 2026 09:18:35 +0000

On Monday, 31 August 2026, at 16:00, two lectures will be held as part of the MONDAY SEMINAR OF COMPUTER SCIENCE AND INFORMATICS organized by the Departments of Information Sciences and Technologies of UP FAMNIT and UP IAM. TIME/LOCATION: 31 August 2026 at 16:00 via Zoom. 1st lecture LECTURER: Matic POŽAR Matic Požar is a first year PhD student and teaching assistant at UP FAMNIT. His research interests include graph representation learning, graph algorithms, and influence maximization. TITLE: A New Approximation Method for the Independent Cascade Influence Function ABSTRACT: The Independent Cascade model is one of the most commonly used diffusion models for representing influence spread in networks. Evaluating this model is crucial for solving the well-known influence maximization problem and its variants; however, this evaluation is #P-hard. Consequently, most implementations rely on approximation methods.Monte Carlo simulations can approximate the independent cascade model with arbitrary accuracy, but only by the cost of a high number of samples, making them impractical for large networks.We propose a new sampling-free approximation algorithm and conduct extensive experiments comparing it to existing approaches. Our results show that the proposed method is among the most accurate while being substantially more efficient. The seminar will be held in English. ============================================================================================================= 2nd lecture: LECTURER: Uroš SERGAŠ Uroš Sergaš is a teaching assistant of Computer Science at the Faculty of Mathematics, Natural Sciences and Information Technologies and is aa member of Centre for Responsible AI UP. As a researcher specializing in recommender systems and computational social science, his work focuses on applying machine learning methods to address societal challenges. Recently, his work also addresses the issue of AI alignment. He is currently pursuing his PhD in recommender systems. TITLE: Prompt to Press: Evaluating Human Perception of AI Involvement in News Writing Across Prompt Specificity ABSTRACT: Large language models (LLMs) are becoming a common feature in content creation tools, prompting important questions about how design choices influence user trust and engagement in AI-assisted journalism. Beyond output quality, factors such as prompt specificity, model choice, and authorship disclosure are themselves interaction design parameters that influence how users interpret and evaluate AI contributions. Yet, little is known about how these design decisions affect reader perceptions in journalistic contexts. To address this gap, we conducted an experiment with 150 participants who evaluated news articles on the sensitive topic of assisted suicide. The articles systematically varied in authorship (human-written, AI-edited, or AI-generated), stance (pro- or anti-legalization), and prompt specificity (vague, moderate, or highly detailed). Participants rated each article on engagement, subjectivity, and perceived AI involvement, and also provided open-ended justifications for their authorship judgments. Our findings show that prompt specificity and model choice significantly influence perceptions of authorship, underscoring how technical design decisions in AI tools can shape public trust in journalism. The seminar will be held in English. ============================================================================================================= The seminars will be held online via the Zoom application, starting at 16:00, at the following link: https://upr-si.zoom.us/j/297328207?pwd=S3Zpdk1VR3pjckNtWkQwKzlvcDR5UT09Meeting ID: 297 328 207Passcode: 123456789 You are cordially invited!

The post Matic POŽAR and Uroš SERGAŠ appeared first on UP Famnit.

Nemanja CVETIĆ and Tilen GIMPELJ

vanja.kovse — Mon, 24 Aug 2026 10:06:17 +0000

On Monday, August 24, 2026, at 4:00 PM, two lectures will be held as part of the MONDAY RIN SEMINAR, organized by the Departments of Information Sciences and Technologies UP FAMNIT and UP IAM. TIME/LOCATION: August 24, 2026, at 4:00 PM, via Zoom. —————————————————————————————————————————————————————- 1st lecture: Lecturer: Nemanja CVETIĆ Nemanja Cvetić is a Master’s student in Computer Science with a Bachelor’s degree in Information Technology. His research interests include machine learning, explainable artificial intelligence (XAI), data science, computer science and software engineering. His current research focuses on evaluating model interpretability techniques and their practical applications in machine learning. Title: Model interpretability in machine learning: A Comparison of LinearSHAP TreeSHAP, and LIME on the California Housing Dataset—————————————————————————————————————————————————————- ABSTRACT: This seminar presents an empirical comparison of SHAP (LinearSHAP and TreeSHAP) and LIME as post-hoc explainability methods for machine learning models. Using Linear Regression and Random Forest models trained on the California Housing dataset, the study evaluates the quality, consistency, and interpretability of feature attributions. The results demonstrate that SHAP provides more stable and theoretically grounded explanations, while LIME shows greater variability, particularly for complex non-linear models. The seminar will be conducted in English. —————————————————————————————————————————————————————- 2nd lecture: Tilen GIMPELJ Tilen Gimpelj is a student completing his 2nd year of the master’s program Data Science at UP FAMNIT. He works at the Department of Applied Natural Sciences under Assoc. Prof. Dr. Marko Jukić, and as a student he works at Novartis as a developer in the Bioprocess Development Department. —————————————————————————————————————————————————————- Title: Comparison of clustering algorithms for identifying conserved waters in an E. coli DNA gyrase ABSTRACT: The seminar compares three algorithms – KMeans, DBSCAN and OPTICS for the purpose of identifying conserved water molecules in the DNA gyrase of E.Coli using water oxygen positions from experimentally determined protein structures aligned to the reference structure 4DUH. The seminar will be conducted in English. The two seminars will be held online via Zoom, starting at 4:00 PM, at the following link: https://upr-si.zoom.us/j/297328207?pwd=S3Zpdk1VR3pjckNtWkQwKzlvcDR5UT09 Meeting ID: 297 328 207Passcode: 123456789 You are cordially invited!

The post Nemanja CVETIĆ and Tilen GIMPELJ appeared first on UP Famnit.

Pika Povh Mavrič and Nemanja Cvetić

vanja.kovse — Mon, 03 Aug 2026 07:51:17 +0000

On Monday, August 3, 2026, at 4:00 PM, two lectures will be held as part of the Monday RIN seminar, organized by the Departments of Information Sciences and Technologies UP FAMNIT and UP IAM. Time/Location: August 3, 2026, at 4:00 PM, via Zoom. Lecture 1:============ —————————————————Lecturer: Pika POVH MAVRIČ————————————————— Pika Povh Mavrič is currently completing her master’s studies in Computer Science at UP FAMNIT. She works as a data analyst in the Customer Relationship Management department at a bank, where she applies and deepens her expertise in data analytics. Her work involves automating processes, improving data quality, creating various reports and analyses, and applying data-driven decision-making to everyday processes within the department. The topic of her paper aligns well with her work in the banking sector. ——————————————————————————————————Title: Machine learning models for predicting personal loan acquisition—————————————————————————————————— Summary: In modern retail banking, the traditional approach of mass marketing consumer loans is becoming highly inefficient and often leads to customer fatigue. To address this issue, this research project focuses on developing a machine learning model designed to accurately predict a customer’s intent to acquire a personal loan. By analyzing historical data, including demographics, product ownership, and past marketing engagement, the project uncovers hidden behavioral patterns. This predictive approach allows financial institutions to shift towards a more customer-centric strategy, optimizing marketing resources and offering loans only to those who truly need them. The seminar will be conducted in English. ============================================================================================================= Lecture 2: ============ ——————————————-Lecturer: Nemanja CVETIĆ——————————————- Nemanja Cvetić is a Master’s student in Computer Science with a Bachelor’s degree in Information Technology. His research interests include machine learning, explainable artificial intelligence (XAI), data science, computer science and software engineering. His current research focuses on evaluating model interpretability techniques and their practical applications in machine learning. ——————————————————————————————————————————————————————————————Title: Model interpretability in machine learning: A Comparison of LinearSHAP TreeSHAP, and LIME on the California Housing Dataset—————————————————————————————————————————————————————————————— Summary: This seminar presents an empirical comparison of SHAP (LinearSHAP and TreeSHAP) and LIME as post-hoc explainability methods for machine learning models. Using Linear Regression and Random Forest models trained on the California Housing dataset, the study evaluates the quality, consistency, and interpretability of feature attributions. The results demonstrate that SHAP provides more stable and theoretically grounded explanations, while LIME shows greater variability, particularly for complex non-linear models. The seminar will be conducted in English. ============================================================================================================= The seminars will be held online via Zoom, starting at 4:00 PM, using the following link: https://upr-si.zoom.us/j/297328207?pwd=S3Zpdk1VR3pjckNtWkQwKzlvcDR5UT09 Meeting ID: 297 328 207Passcode: 123456789 You are kindly invited!

The post Pika Povh Mavrič and Nemanja Cvetić appeared first on UP Famnit.

Dušan Todorović

vanja.kovse — Mon, 27 Jul 2026 07:37:07 +0000

On Monday, July 27, 2026, at 4:00 PM, a lecture will be held as part of the Monday RIN seminar, organized by the UP FAMNIT and UP IAM. TIME / VENUE: July 27, 2026, at 4:00 PM, via Zoom. SPEAKER: Dušan TODOROVIĆ Dušan Todorović graduated from the Academy of Technical and Educational Applied Studies (ATVSS) in Niš, specializing in Modern Computer Technologies. He is currently a master’s student in the Computer Science program at UP FAMNIT. His seminar paper was prepared as part of the Research Seminar course under the supervision of Dr. Branko Kavšek. ———————————————————————————————————NASLOV: Tehnike za razložljivost v strojnem učenju: LIME, SHAP in Grad-CAM——————————————————————————————————— POVZETEK: V seminarski nalogi obravnavam problem interpretabilnosti kompleksnih modelov strojnega učenja ter analiziram tri pomembne metode za razložljivost — LIME, SHAP in Grad-CAM. V teoretičnem delu predstavim osnovne pojme interpretabilnosti (lokalna in globalna interpretacija, model-agnostične in model-specifične metode) ter matematično ozadje posamezne metode. V praktičnem delu izvedem dva eksperimenta: prvega na podatkovnem naboru Titanic z modelom Random Forest (LIME in SHAP za tabelarične podatke), drugega pa na podatkovnem naboru Fashion-MNIST s konvolucijsko nevronsko mrežo (SHAP in Grad-CAM za slike). Rezultati pokažejo, da LIME ponuja intuitivne lokalne razlage, SHAP omogoča stabilnejšo in teoretično bolje utemeljeno interpretacijo (tako lokalno kot globalno), Grad-CAM pa vizualni vpogled v regije slike, ki najbolj vplivajo na odločitev CNN modela. Ugotovim, da ne obstaja ena univerzalna metoda za razlago vseh modelov, saj je izbira odvisna od tipa podatkov in modela. Seminar bo potekal v slovenskem jeziku. Seminar bo potekal online prek aplikacije Zoom s pričetkom ob 16:00 uri na sledeči povezavi: https://upr-si.zoom.us/j/297328207?pwd=S3Zpdk1VR3pjckNtWkQwKzlvcDR5UT09 Meeting ID: 297 328 207Passcode: 123456789 Vabljeni!

The post Dušan Todorović appeared first on UP Famnit.

Lina MEŠIČ in Luka MATIČ

mario.krajoski — Sun, 28 Jun 2026 22:00:00 +0000

V ponedeljek, 29. junija 2026, bodo ob 16:00 uri izvedeni dve predavanji v okviru PONEDELJKOVEGA SEMINARJA RAČUNALNIŠTVA IN INFORMATIKE Oddelkov za Informacijske znanosti in tehnologije UP FAMNIT in UP IAM. ČAS/PROSTOR: 29. junij 2026 ob 16.00 prek Zoom-a. 1. predavanje: ============ —————————————– PREDAVATELJICA: Lina MEŠIČ —————————————– Lina Mešič je magistrska študentka Računalništva in informatike na UP FAMNIT in trener športnega plezanja. ———————————————————————————————– NASLOV: Interaktivni sistemi kot orodje za trening športnega plezanja ———————————————————————————————– POVZETEK: Športno plezanje je v zadnjem desetletju doživelo izjemen razvoj, s tem pa so v ospredje stopile določene pomanjkljivosti v treniranju in povečale zahteve po kakovostnem treningu. V seminarju bo predstavljen pregled devetih interaktivnih sistemov s področja HCI, ki se uporabljajo za trening športnega plezanja. Od plezalnih sten s projiciranimi igrami, kjer plezalci lovijo navidezne cilje in pozabijo na strah pred višino, prek virtualne resničnosti za varno soočanje s padci, do nosljivih naprav, ki prek vibracij ali zvoka vodijo plezalca med samim vzponom. Na podlagi izbranih študij poskušamo odgovoriti na vprašanja, kje so meje teh sistemov, komu najbolj koristijo in kako jih v prihodnosti oblikovati, da bodo trenerju v pomoč in ne v napoto. Ugotovitve kažejo, da tehnologija seveda ne more nadomestiti trenerja, lahko pa prevzame ponavljajoče se naloge in administracijo ter mu tako olajša delo. Seminar bo potekal v slovenskem jeziku. =============================================================================== 2. predavanje: ============ ————————————- PREDAVATELJ: Luka MATIČ ————————————- Luka Matič je študent magistrskega študija Računalništva in informatike na UP FAMNIT, ki se osredotoča na biomedicinsko obdelavo signalov in razvoj programske opreme. Njegova raziskovalna področja vključujejo analizo EEG signalov, strojno učenje in obdelavo podatkov v realnem času. Trenutno se ukvarja s primerjavo spektralnih in kepstralnih značilk pri klasifikaciji EEG signalov. ——————————————————————————————– NASLOV: Uporaba kepstralne analize za klasifikacijo EEG posnetkov ——————————————————————————————– POVZETEK: Seminar obravnava primerjavo spektralnih in kepstralnih metod za analizo in klasifikacijo EEG signalov. Predstavljena je ekstrakcija značilk iz EEG podatkov v frekvenčni in kepstralni domeni ter njihova uporaba v okviru metod strojnega učenja. Osrednji poudarek je na primerjalni analizi učinkovitosti obeh pristopov, ki temelji na doseženi točnosti in pravilnosti klasifikacije EEG signalov. Rezultati prispevajo k boljšemu razumevanju vpliva izbire metode na uspešnost klasifikacije EEG signalov. Seminar bo potekal v slovenskem jeziku. =============================================================================== Seminarja bosta potekala online prek aplikacije Zoom s pričetkom ob 16:00 uri na sledeči povezavi: https://upr-si.zoom.us/j/297328207?pwd=S3Zpdk1VR3pjckNtWkQwKzlvcDR5UT09 Meeting ID: 297 328 207 Passcode: 123456789 Vabljeni!

The post Lina MEŠIČ in Luka MATIČ appeared first on UP Famnit.

Jovan VUKOVIĆ in Uroš SERGAŠ

mario.krajoski — Sun, 21 Jun 2026 22:00:00 +0000

V ponedeljek, 22. junija 2026, bodo ob 16:00 uri izvedeni dve predavanji v okviru PONEDELJKOVEGA SEMINARJA RAČUNALNIŠTVA IN INFORMATIKE Oddelkov za Informacijske znanosti in tehnologije UP FAMNIT in UP IAM. ČAS/PROSTOR: 22. junij 2026 ob 16.00 prek Zoom-a. 1. predavanje: ============ —————————————– PREDAVATELJ: Jovan VUKOVIĆ —————————————– Jovan Vuković is a Master's student in Data Science at the University of Primorska (UP FAMNIT). He previously completed a Bachelor's degree in Computing and Information Technology at the University of Montenegro. His professional background is primarily focused on software engineering, microservice architectures, distributed systems, and large-scale backend development. His current research interests include knowledge graphs, machine learning, data integration, and scalable graph processing. ————————————————————————————————————————————– NASLOV: Embedding-Based Entity Alignment between Large-Scale Knowledge Graphs: A Case Study of YAGO and OpenAlex ————————————————————————————————————————————– POVZETEK: Knowledge graphs have become an important approach for representing and integrating structured information in domains such as web search, digital libraries, recommendation systems, and scientific knowledge management. By representing entities and their relationships as graphs, they enable the discovery of connections that are difficult to capture using traditional relational data models. As the number and size of available knowledge graphs continue to grow, combining information from multiple sources has become an increasingly important challenge. One of the key problems in this context is entity alignment, the task of identifying entities in different knowledge graphs that refer to the same real-world object. Solving this problem is essential for building integrated knowledge bases, but it becomes particularly challenging when working with large-scale graphs containing hundreds of millions of entities and relationships. This presentation introduces the entity alignment problem through a case study involving the YAGO and OpenAlex knowledge graphs. It focuses on the first stages of a scalable alignment pipeline, including large-scale RDF preprocessing, extraction of structural and textual information, graph transformation and integer encoding, and the generation of knowledge graph embeddings using PyTorch-BigGraph. Several embedding models, including TransE, DistMult, and ComplEx, are evaluated through link prediction experiments on large-scale datasets. The obtained results are used to assess how well the learned embeddings capture the underlying graph structure and to identify the most suitable embedding model for the subsequent entity alignment task. Seminar bo potekal v angleškem jeziku. =============================================================================== 2. predavanje: ============ ————————————– PREDAVATELJ: Uroš SERGAŠ ————————————– Uroš Sergaš is a teaching assistant of Computer Science at UP FAMNIT and a member of the Centre for Responsible AI UP. As a researcher specializing in recommender systems and computational social science, his work focuses on applying machine learning methods to address societal challenges. Recently, his work also addresses the issue of AI alignment. He is currently pursuing his PhD in recommender systems. ————————————————————————————————————————————– NASLOV: Can News Ranking Reduce Political Polarization? Evidence from an Online Experiment with LLM-Generated News ————————————————————————————————————————————– POVZETEK: Algorithmic news ranking and large language models (LLMs) increasingly mediate how citizens receive political information. While personalization is often criticized for reinforcing echo chambers, ranking has also been proposed as a lever for depolarization by shaping exposure to cross-cutting viewpoints. We tested this claim in an online experiment (N=100, Prolific) using five LLM-generated news articles on gun legislation spanning pro–gun freedom to pro–gun control. All participants read the same articles; the only manipulation was article order, instantiated as a random baseline and six stance-aware depolarization-oriented strategies (counter-narrative sandwich, balanced alternation, and directional gradients). Pre–post questionnaires measured ideological selfplacement (feeling thermometer) and affective evaluations of gun-control and gun-freedom advocates. Across six research questions, we tested to see whether such recommender strategies had any reduction in ideological or affective polarization. Seminar bo potekal v angleškem jeziku. =============================================================================== Seminarja bosta potekala online prek aplikacije Zoom s pričetkom ob 16:00 uri na sledeči povezavi: https://upr-si.zoom.us/j/297328207?pwd=S3Zpdk1VR3pjckNtWkQwKzlvcDR5UT09 Meeting ID: 297 328 207 Passcode: 123456789 Vabljeni!

The post Jovan VUKOVIĆ in Uroš SERGAŠ appeared first on UP Famnit.

Ekaterina BOCHVAROSKA: RAVEN: A Retention-Aware Verifiable Extension-Block Network for Off-Chain Message Storage and Retrieval in Andrej Milisavljević: SAVI ionization and conformer generation

mario.krajoski — Sun, 14 Jun 2026 22:00:00 +0000

V ponedeljek, 15. junija 2026, bodo ob 16:00 uri izvedeni dve predavanji v okviru PONEDELJKOVEGA SEMINARJA RAČUNALNIŠTVA IN INFORMATIKE Oddelkov za Informacijske znanosti in tehnologije UP FAMNIT in UP IAM. ČAS/PROSTOR: 15. junij 2026 ob 16.00 prek Zoom-a. 1. predavanje: ============ ———————————————————- PREDAVATELJICA: Ekaterina BOCHVAROSKA ———————————————————- Ekaterina Bochvaroska received her B.Sc. degree in Computer Science from UP FAMNIT in 2025, where she is currently pursuing the M.Sc. degree in Computer Science. She is currently working as a developer at the Port of Koper, where she is involved in software development, data analysis, and digital solutions for logistics and port operations. Her research interests include decentralized systems, distributed computing, blockchain technologies, algorithms and data structures. ————————————————————————————————————————————- NASLOV: RAVEN: A Retention-Aware Verifiable Extension-Block Network for Off-Chain Message Storage and Retrieval ————————————————————————————————————————————- POVZETEK: RAVEN is a prototype protocol and reference implementation that stores encrypted-message data in off-chain extension blocks and commits each block to an ordered parent chain through a small anchor record. Nodes keep different amounts of history – archive nodes retain everything, regular nodes only a recent suffix window and a client retrieves height ranges from whichever nodes can serve them. Every returned block is verified against its canonical, final anchor before acceptance, so corrupt, missing, or orphaned data is detected rather than trusted. The project covers the protocol core, the node runtime, the retention-aware retrieval pipeline, and an experiment that runs the same workload across several strategies on a local Docker Compose demo. Seminar bo potekal v angleškem jeziku. ============================================================================== 2. predavanje: ============ ————————————————– PREDAVATELJ: Andrej MILISAVLJEVIĆ ————————————————– Andrej Milisavljević is a student of Data Science at UP FAMNIT, currently researching cheminformatics at UPR with professors Marko Jukić and Jure Pražnikar as mentors. Their research includes methods for protein modeling, binding site detection, ligand docking, and the parallelization and distribution of cheminformatics computational analyses. ———————————————————————- NASLOV: SAVI ionization and conformer generation ———————————————————————- POVZETEK: In molecular docking simulations used for the discovery of new medicines, the initial conformation and ionization states of the molecules being studied are crucial for ensuring biochemically relevant results and allowing drug-like molecules to be recognized. In this work, we describe a distributed and parallelized pipeline for efficiently computing ionization states and molecular conformations of 1 billion molecules from the Synthetically Accessible Virtual Inventory (SAVI), and the resulting docking-ready dataset of drug-like molecules for use in drug discovery and other related studies. Seminar bo potekal v angleškem jeziku. ============================================================================== Seminarja bosta potekala online prek aplikacije Zoom s pričetkom ob 16:00 uri na sledeči povezavi: https://upr-si.zoom.us/j/297328207?pwd=S3Zpdk1VR3pjckNtWkQwKzlvcDR5UT09 Meeting ID: 297 328 207 Passcode: 123456789 Vabljeni!

The post Ekaterina BOCHVAROSKA: RAVEN: A Retention-Aware Verifiable Extension-Block Network for Off-Chain Message Storage and Retrieval in Andrej Milisavljević: SAVI ionization and conformer generation appeared first on UP Famnit.