Intelligent Data Science &
Applications Lab, Korea University

We develop intelligent systems that help people navigate overwhelming information. We deliver relevant, high-quality, and actionable insights for each user’s context and needs.

IDEA Lab members at DASFAA

Recent News

Accepted PaperEMNLP 2026 Demo

hoBIT: A Profile-Aware Retrieval-Augmented Chatbot for University Academic Advising

Accepted PaperEMNLP 2026

PEARL: Front-Loading Relational Chains for Multi-Hop Table Retrieval

Accepted PaperCIKM 2026

Beyond Observed Auxiliary Relations: Environment-Conditioned Modeling for Multi-Behavior Recommendation

Accepted PaperCIKM 2026

SCoRD: Semantic-Assisted Continual Retriever-Reranker Distillation for LLM-Based Recommendation

Accepted PaperCIKM 2026

Profiling What Matters: Context-Aware Item Profiles from Large-Scale Metadata for LLM Recommenders

Accepted PaperCIKM 2026

Dynamic Spectral Decoupling for Time-Series Forecasting

Accepted PaperCIKM 2026

TRACER: Balancing Stability-Plasticity-Cognitivity Trilemma for LLM Enhanced Continual Recommendation

Accepted PaperCOLM 2026

Uncertainty-Aware Variational Reward Factorization via Probabilistic Preference Bases for LLM Personalization

Accepted PaperTIST 2026

Improving Scientific Document Retrieval with Academic Concept Index

Award2026

Prof. Kang received the Seoktop Lecture Award (Big Data Analytics) from Korea University.

Accepted PaperKDD 2026

CoRank: LLM-Based Compact Reranking with Document Features for Scientific Retrieval

Talk2026

Invited talk: “Knowledge-Structured Retrieval for Scientific Literature”, Data Intelligence Workshop, 2026.

Accepted PaperSIGIR 2026

SPRINT: Scalable and Predictive Intent Refinement for LLM-Enhanced Session-based Recommendation

Accepted PaperSIGIR 2026

Filling the Gaps: Selective Knowledge Augmentation for LLM Recommenders

Accepted PaperSIGIR 2026

FLAME: Condensing Ensemble Diversity into a Single Network for Efficient Sequential Recommendation

Accepted PaperSIGIR 2026

MVIGER: Multi-View Variational Integration of Complementary Knowledge for Generative Recommender

Accepted PaperACL 2026 Findings

Why These Documents? Explainable Generative Retrieval with Hierarchical Category Paths

Talk2026

Invited talk: “Knowledge-Structured Retrieval for Scientific Literature”, CSE/GSAI Seminar, POSTECH, 2026.

Accepted PaperICLR 2026

Continual Low-Rank Adapters for LLM-based Generative Recommender Systems

TutorialDASFAA 2026

Continual Recommender Systems: A Focus on LLMs and Evolving Trends

Accepted PaperWSDM 2026

Capturing User Interests from Data Streams for Continual Sequential Recommendation

Accepted PaperWWW 2026

PairSem: LLM-Guided Pairwise Semantic Matching for Scientific Retrieval

Accepted PaperKDD 2026

CREAM: Continual Retrieval on Dynamic Streaming Corpora with Adaptive Soft Memory

Accepted PaperTKDE 2026

BPL: Bias-adaptive Preference Distillation Learning for Recommender System

Service2026

Prof. Kang was invited as an Area Chair of KDD and the EMNLP System Demonstration Track, and to program committees including AAAI, SIGIR, NeurIPS, CIKM, ICDM, SIGIR-AP, and ACM AI Summit.

Scholarship2025

Seunghan was selected as a recipient of the AI SeoulTech Scholarship!

Award2025

Prof. Kang received the Seoktop Lecture Award (Artificial Intelligence) from Korea University.

Service2025

Prof. Kang was invited as a PhD Mentor for the CIKM 2025 PhD Symposium.

Award2025

Our scientific document retrieval paper was selected as one of the Best Papers of WSDM 2025 and invited to ACM TIST.

Accepted PaperEMNLP 2025

Topic Coverage-based Demonstration Retrieval for In-Context Learning

Accepted PaperEMNLP 2025 Findings

Scientific Paper Retrieval with LLM-Guided Semantic-Based Ranking

TutorialCIKM 2025

Continual Recommender Systems

Accepted PaperCOLM 2025

DeepRetrieval: Hacking Real Search Engines and Retrievers with Large Language Models via Reinforcement Learning

Accepted PaperCOLM 2025

Imagine All The Relevance: Scenario-Profiled Indexing with Knowledge Expansion for Dense Retrieval

Accepted PaperKDD 2025

Delving into Instance-Dependent Label Noise in Graph Data: A Comprehensive Study and Benchmark

Accepted PaperSIGIR 2025

Embracing Plasticity: Balancing Stability and Plasticity in Continual Recommender Systems

Accepted PaperSIGIR 2025

Personalized Preference Reasoning with Large Language Models for Accurate and Explainable Recommendation

Accepted PaperWWW 2025

Uncertainty Quantification and Decomposition for LLM-based Recommendation

Accepted PaperWWW 2025

Chain-of-Factors Paper-Reviewer Matching

Accepted PaperWSDM 2025

Improving Scientific Document Retrieval with Concept Coverage-based Query Set Generation

Accepted PaperWSDM 2025

Unsupervised Robust Cross-Lingual Entity Alignment via Neighbor Triple Matching with Entity and Relation Texts

Service2025

Prof. Kang was invited to program committees including KDD, WWW, AAAI, SIGIR, SIGIR-AP, ACL, AACL-SRW, CVPR, and DASFAA.

Research Areas

We explore three connected research directions for intelligent information systems.

Recommender Systems

Personalization
User Simulation
LLM-based Ranking
Continual Recommendation
Multimodal Learning

Information Retrieval

Agentic Search
Graph RAG
Ontology Construction
Scientific Retrieval
RAG & Grounding
Multimodal Learning

Data & Web Mining

Web-scale Data Mining
User Behavior Mining
Multimodal Learning
Temporal Data Mining
Knowledge Structures