Publications

Our group has the following papers accepted at EMNLP’26:

• A Survey on Natural Language User Profiles for Recommendation: Methods, Datasets, and Metrics – main conference paper by Mariam Arustashvili and Krisztian Balog [PDF]
• “Act Like a 5th Grader” is Not Enough: Bounding Knowledge in LLM-Based User Simulators – findings paper by Krisztian Balog and Arild Michel Bakken [PDF]
• Does the Greater-Than Circuit Generalize? A Reproducibility Study Across Models and Numerical Entailment – BlackboxNLP workshop paper by Vinay Setty

Our group has the following papers accepted at SIGIR’26:

• A Standardized Re-evaluation of Conversational Recommender Systems on the ReDial Dataset – reproducibility paper by Ivica Kostric and Krisztian Balog [PDF]
• A Reproducibility Study of Metacognitive Retrieval-Augmented Generation – reproducibility paper by Gabriel Iturra-Bocaz and Petra Galuscakova [PDF]

Additionally, our group has two papers accepted at ICTIR’26:

• RecQuest: Towards Estimating User Domain Knowledge in Conversational Recommender Systems – research article by Ivica Kostric, Ujwal Gadiraju and Krisztian Balog [PDF]
• Effectiveness of Sparse Retrieval for Automated Fact-Checking – research article by Ritvik Setty and Vinay Setty [PDF]

The paper Should We Tailor the Talk? Understanding the Impact of Conversational Styles on Preference Elicitation in Conversational Recommender Systems by Ivica Kostric, Krisztian Balog, and Ujwal Gadiraju has been accepted at the 33rd ACM International Conference on User Modeling, Adaptation and Personalization (UMAP ’25) [PDF].

Our group has the following papers accepted at SIGIR’25:

• GINGER: Grounded Information Nugget-Based Generation of Responses — short paper by Weronika Łajewska and Krisztian Balog [PDF]
• MultiConAD: A Unified Multilingual Conversational Dataset for Early Alzheimer’s Detection — resource paper by Arezo Shakeri, Mina Farmanbar, Krisztian Balog [PDF]
• Navigating Speech Recording Collections with AI-Generated Illustrations — demo paper by Sirina Håland, Trond Karlsen Strøm and Petra Galuščáková

In addition, our group is also involved with the organization of

• Simulations for Information Access (Sim4IA) workshop, featuring two micro shared tasks
• Theory and Toolkits for User Simulation in the Era of Generative AI: User Modeling, Synthetic Data Generation, and System Evaluation tutorial (website)

CRS Arena is a platform to benchmark conversational recommender systems in a crowdsourced environment, developed by Nolwenn Bernard, Hideaki Joko, Faegheh Hasibi, and Krisztian Balog. This work has been presented at the 18th ACM International Conference on Web Search and Data Mining (WSDM ’25), in Hannover. Try it out here or have a look at the demo paper.

Some recent paper acceptances on fact checking:

• LiveFC: A System for Live Fact-Checking of Audio Streams — short paper at WSDM ’25 by V. Venktesh and Vinay Setty [PDF]
• FlashCheck: Exploration of Efficient Evidence Retrieval for Fast Fact-Checking — “IR for Good” paper at ECIR ’25 by Kevin Nanhekhan, V. Venktesh, Erik Martin, Henrik Vatndal, Vinay Setty and Avishek Anand [PDF]
• Annotation Tool and Dataset for Fact-Checking Podcasts — resource paper at WWW ’25 by Vinay Setty and Adam James Becker [PDF]
• FactIR: A Real-World Zero-shot Open-Domain Retrieval Benchmark for Fact-Checking — resource paper at WWW ’25 by Venktesh V and Vinay Setty [PDF]

The paper Can Users Detect Biases or Factual Errors in Generated Responses in Conversational Information-Seeking? by Weronika Łajewska, Krisztian Balog, Damiano Spina, and Johanne Trippas has been accepted at the 2nd International ACM SIGIR Conference on Information Retrieval in the Asia Pacific (SIGIR-AP ’24). The preprint will follow soon.

QuestGen: Effectiveness of Question Generation Methods for Fact-Checking Applications by Ritvik Setty and Vinay Setty has been accepted at the 33rd ACM International Conference on Information and Knowledge Management (CIKM’24) as a short paper.

Some recent paper acceptances on user simulation in conversational information access:

• Identifying Breakdowns in Conversational Recommender Systems using User Simulation — full paper at CUI ’24 by Nolwenn Bernard and Krisztian Balog [PDF]
• Towards a Formal Characterization of User Simulation Objectives in Conversational Information Access — full paper at ICTIR ’24 by Nolwenn Bernard and Krisztian Balog [PDF]

The book User Simulation for Evaluating Information Access Systems by Krisztian Balog and ChengXiang Zhai has been published by Foundations and Trends in Information Retrieval. This comprehensive monograph delves into the pivotal role of user simulation in assessing the effectiveness of information access systems, such as search engines, recommender systems, and conversational assistants. A preprint is available on arXiv.