Publications

The paper Generating Usage-related Questions for Preference Elicitation in Conversational Recommender Systems by Ivica Kostric, Krisztian Balog, and Filip Radlinski has been accepted by ACM Transactions on Recommender Systems and is now available online.

Our group has two contributions accepted at CIKM’23:

• Towards Filling the Gap in Conversational Search: From Passage Retrieval to Conversational Response Generation — resource paper by Weronika Łajewska and Krisztian Balog [PDF]
• Tutorial on User Simulation for Evaluating Information Access Systems — tutorial by Krisztian Balog and ChengXiang Zhai [website]

Trustworthy journalism through AI, co-authored by Vinay Setty, has been accepted in the Data & Knowledge Engineering journal and is now available here.

MG-ShopDial: A Multi-Goal Conversational Dataset for e-Commerce by Nolwenn Bernard and Krisztian Balog has been accepted at the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR’23) as a resource paper.

Extreme Classification for Answer Type Prediction in Question Answering by Vinay Setty has been accepted at Joint Conference on Digital Libraries (JCDL’23) as a short paper.

UserSimCRS is an extensible user simulation toolkit for evaluating conversational recommender systems developed by Jafar Afzali, Krisztian Balog, Nolwenn Bernard, Aleksander Drzewiecki, and Shuo Zhang. This work has been presented at the 16th ACM International Conference on Web Search and Data Mining (WSDM ’23), in Singapore. Have a look at the demo paper or the toolkit for more details.

The paper describing our participation at the TREC 2022 Conversational Assistance Track is now available here.

The paper From Baseline to Top Performer: A Reproducibility Study of Approaches at the 2021 TREC Conversational Assistance Track by Weronika Łajewska and Krisztian Balog has been accepted at the 45th European Conference on Information Retrieval (ECIR’23). A preprint is available here.

DAFiNN is a conversational conference assistant that has been developed by members of our group, Ivica Kostric, Krisztian Balog, Nolwenn Bernard, Weronika Łajewska, together with students at the University of Stavanger. This work has been demonstrated at the 16th ACM Conference on Recommender Systems (RecSys‘22) last week, in Seattle, USA. Have a look at the dagfinn.ai website, the demo paper, or the video with a walk-through of its capabilities.

Our group has three papers accepted at SIGIR’22:

• Analyzing and Simulating User Utterance Reformulation in Conversational Recommender Systems — full paper by former IAI member Shuo Zhang, Mu Chun Wang, and Krisztian Balog [PDF]
• Would You Ask it that Way? Measuring and Improving Question Naturalness for Knowledge Graph Question Answering — resource paper by Trond Linjordet and Krisztian Balog [PDF]
• SparCAssist: A Model Risk Assessment Assistant Based on Sparse Generated Counterfactuals — demo paper by Zijian Zhang, Vinay Setty, and Avishek Anand [PDF]