Data Mining Lab.

Welcome to Data Mining Laboratory in the Department of Computer Science and Engineering at Seoul National University. Our research interests lie in artificial intelligence (AI), data mining, and machine learning to find models, algorithms, and systems for data analysis. Specifically, we focus on the following research topics: deep learning & machine learning, recommendation system, graphs/tensors, and financial AI.

What's New

  • [Jul. 2024] A paper accepted to CIKM 2024, a top tier data mining conference.

    A paper is accepted to CIKM 2024, a top tier data mining conference. The paper "Fast and Accurate PARAFAC2 Decomposition for Time Range Queries on Irregular Tensors" proposed REPEAT, a fast and accurate PARAFAC2 decomposition method for handling arbitrary time range queries on irregular tensors.

  • [Jun. 2024] A paper accepted to Interspeech 2024, a top tier speech and language processing conference.

    A paper accepted to Interspeech 2024, a top tier speech and language processing conference. The paper "Domain-Aware Data Selection for Speech Classification via Meta-Reweighting" proposed DoReMe, a data selection method for accurate speech classification on a target domain.

  • [May. 2024] 3 papers accepted to KDD 2024, the world-best data science and AI conference.

    3 papers from Data Mining Lab are accepted to KDD 2024, the world-best data science and AI conference. Congratulations!

  • [Feb. 2024] A paper accepted to PAKDD 2024, a top tier data mining conference.

    A paper is accepted to PAKDD 2024, a top tier data mining conference. The paper “Accurate Semi-supervised Automatic Speech Recognition via Multi-hypotheses-based Curriculum Learning” proposed an accurate semi-supervised automatic speech recognition framework utilizing multiple hypotheses of speech instances and curriculum learning to consider uncertainty of pseudo labels and improve speech recognition performance.

  • [Jan. 2024] A paper accepted to The Web Conference 2024, a top tier data mining conference.

    A paper is accepted to The Web Conference 2024, a top tier data mining conference. The paper "Accurate Cold-start Bundle Recommendation via Popularity-based Coalescence and Curriculum Heating" proposed COHEAT, an accurate method for cold-start bundle recommendation.

<< Older news