Location

To be held at the International Multiconference on Information Society IS-2026

Conference themes

  • Statistical Data Analysis and Causal Inference
  • Data, Text, Graph, Web, and Multimedia Mining
  • Large Language Models and Generative AI
  • Link Detection, Social Network Analysis, and Knowledge Graphs
  • Natural Language Processing and Semantic Technologies
  • Sensor Data Analytics, Stream Mining, and Edge AI
  • Explainable AI (XAI), Ethics, and Trustworthy Machine Learning
  • Modeling Complex Systems and AI for Science
  • Applications

Instructions for Authors

Authors are invited to contribute papers that present the state-of-the-art, development goals, applications and future trends in the above mentioned areas.

Authors are invited to submit complete papers not exceeding 4 pages, written in English and formatted following the conference formatting instructions.

The authors should submit the papers by August 31, 2026 in .pdf format via submission page, select track on DATA MINING AND DATA WAREHAUSES – SIKDD 2026.In case of any issues let us know at dunja.mladenic@ijs.si with subject "SiKDD paper submissions".

Conference Program

Presentations of accepted papers will be available at @videolectures.net.

9:00 - 11:00
Exploring Feedback-Aware Retrieval for IT Support First-Reply Generation
Domen Jeršek, Blaž Kažič and Klemen Kenda
The Changing Slovenian Smellscape, 1870–1919: Measuring Olfactory Discourse
Inna Novalija, Janez Brank, Dunja Mladenić and Marko Grobelnik
Explaining Headline A/B Test Outcomes with Sparse Autoencoder Features
Drejc Pesjak, Dunja Mladenić and Jan Rupnik
Same Labels, Smaller Bill: An Ablation Study of LLM Claim Classification in Cross-Source News
Georgi Trajkov, Blaž Škrlj, Marko Grobelnik and Dunja Mladenić
A Hybrid Lexicon-LLM Approach to Skill Extraction from Job Advertisements
Natalija Mečkaroska and Georgi Trajkov
Benchmarking Machine-Learning Forecasts of Low-Magnitude Seismicity in the Italy–Slovenia Border Region
Oleksandra Topal, Erik Calcina, Luciano Telesca and Dunja Mladenić
The Effect of Label Only vs. Label Plus Description: Comparing GLiNER1 and GLiNER2 for Clinical Entity Recognition Across Five Languages
Abdul Sittar, Erik Novak and Erik Calcina
Data Efficiency of GLiNER: A Multilingual and Multi-Scale Learning Curve Study
Nina Kokalj, Erik Calcina, Abdul Sittar and Erik Novak
11:00 - 11:30
Break
11:30 - 12:15
Comparing Pseudo-Labeling Criteria for Active Learning on a Service-Desk Dataset
Rok Klančič and Klemen Kenda
LLM-Generated Song Lessons: Perceived Usefulness and Shared-Cache Cost Analysis
Georgi Trajkov
Reliable but Not Right: Automated Evaluation of Generated R&D Ideas
Adrian Mladenić Grobelnik
12:15 - 13:25
Graphs of n-grams for Text Generation and Characteristic Phrase Extraction in the Domain of Smell
Nadezhda Komarova and Inna Novalija
QFS-Composer: Query-Focused Summarization Pipeline for Less Resourced Languages
Vuk Đuranović and Marko Robnik-Šikonja
Operationalising Active Learning for Power Grid Security Assessment: A Human-in-the-Loop Retraining Service for a Digital Twin Platform
Gašper Leskovec, Costas Mylonas and Klemen Kenda
Beyond Named Citations: A Corpus of Resolved Cross-References in EU Legislation
Jakob Vokač, Marko Grobelnik and Dunja Mladenić
Towards Scientific Contradiction Mining: Topic Discovery and Claim Extraction from AI Literature
Andrej Natev, Elham Motamedi and Inna Novalija
Chronos-2 for Spacecraft Telemetry Forecasting
Žiga Klun, Dragi Kocev and Jure Brence
Biomedical Relation Extraction: What Lies Between the Entities?
Nadiia Shpanko, Erik Calcina and Erik Novak
13:25 - 14:00
Lunch break
14:00 - 15:30
E3 Atlas: An Autonomous LLM Agent for Curating and Serving a Departmental Knowledge Base
Drejc Pesjak, Georgi Trajkov, Luis Rei and Dunja Mladenić
Evaluating Large Language Model-Generated Explanations for Defective-Part Prediction in Injection Moulding Using Feature-Attribution Evidence
Seyed Iman Hosseini, Lorenzo Seghesio, Konstantinos Gryllias and Dunja Mladenić
Learning Molecular Representations from SMILES for Acute Toxicity Prediction
Simona Cholakova, Domen Vake and Aleksandar Tošić
Logical Anomaly Detection Using Segmentation Transformers
Yunfei Liang, Matic Fučka and Danijel Skočaj
The Role of Artificial Intelligence in Early Moisture Risk Detection for Wooden Buildings
Kristina Piiarska, Aleksandar Tošić and Pau Brunet
Posterior Contraction Rate and Roadmap to Semiparametric Bernstein-von Mises Theorem for Bayesian Causal Forests
Andrej Srakar
Validation of Proper Noun Lexemes in the Digital Dictionary Database of Slovene Using Wikidata and Large Language Models
Jaka Čibej
Comparative Analysis of Feature Selection and Imbalance Handling for Semiconductor Defect Detection on the SECOM Dataset
Sidorela Uku and Klemen Kenda
15:30 - 16:00
Break
16:00 - 17:15
From Quantum Machine Learning to Quantum Artificial Intelligence: A Systematic Meta-Review
Aleksandar Brodschneider and Matjaž Gams
From Explainability to Accountability: XAI and Fundamental Rights Impact Assessments under the EU AI Act
Lovro Bobnar
Influencing Sentiment in LLM Output via Token Selection Manipulation
Kaja Praprotnik, Niki Hrovatin, Jernej Vičič and Domen Vake
From Label Agreement to Downstream Performance: Evaluating LLM-Generated Training Data
Dario Mammana and Petra Kralj Novak
Short Term Time-Series Forecasting Framework
Gašper Kreft, Aljaž Osojnik, Martin Žnidaršič and Bernard Ženko
Short-Term Load Forecasting at the Device Level: Are Specialized Deep-Learning Approaches Worth the Effort?
Gal Zmazek, Aljaž Osojnik, Martin Žnidaršič and Bernard Ženko
Data-Driven and Model-Based Causal Discovery in Sparse Multivariate Time Series: A Comparative Framework for Sustainable Development Systems
Joao Pita Costa, Andrej Srakar, Oleksandra Topal and Jan Šturm
Legal Categories from Incident Reports: Evaluating 14 LLMs under the EU AI Act
Stefan Kofilovski and Georgi Trajkov
Balancing Sustainable Development Goals through Orchestrating Autonomous Multi-Agent Architecture
Sivaramakrishnan R Guruvayur, Krishnasree Achuthan and Joao Pita Costa
Accurate but Uninformative: Checking an LLM Assessment Pipeline Without Ground Truth
Martin Macak and Marko Grobelnik
Combining Supervised and Unsupervised Learning on SMART Data for Hard Drive Failure Assessment
Luka Petek and Vili Podgorelec
Advertising Is Not Noise: Sponsored Speech Outranks New Terminology in Automatic Term Extraction
Martin Macak

Important dates

  • Submission deadline: 31 August, 2026
  • Notification to authors: 20 September, 2026
  • Deadline for camera-ready papers: 25 September, 2026

Organization

Program chairs

Marko Grobelnik

marko.grobelnik@ijs.si

J.Stefan Institute,
Jamova 39,
1000 Ljubljana,
Slovenia

Dunja Mladenić

dunja.mladenic@ijs.si

J.Stefan Institute,
Jamova 39,
1000 Ljubljana,
Slovenia

Previous events

2025
2024
2023
2022
2021
2020
2019
2018
2017
2016
2015
2014
2013
2012
2011
2010
2009
2008
2007
2006
2005
2004
2002
2001
2000
1999