Papers by Kang-Min Kim
CPR-RAG: Clinical Prior-Regularized Retrieval for Anatomy-Aware 3D CT Report Generation (2026.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to grounding radiology reports from 3D volumetric data are limited due to visual-semantic ambiguity and lack of "normal" context. |
| Approach: | They propose a model-agnostic retrieval-augmented generation framework that integrates clinical priors into the retrieval process. |
| Outcome: | The proposed model improves clinical efficacy across state-of-the-art models. |
“Why do I feel offended?” - Korean Dataset for Offensive Language Identification (2023.findings-eacl)
Copied to clipboard
| Challenge: | Existing methods for detecting offensive content rely on labeled datasets, but few consider low-resource languages with relatively less data available for training. |
| Approach: | They propose to use Korean as a dataset for offensive language identification . they propose to perform abusive language detection and sentiment analysis to help identify offensive languages. |
| Outcome: | The proposed datasets improve the performance of offensive language identification in Korean, while the existing methods are limited. |
Don’t Generate, Classify! Low-Latency Prompt Optimization with Structured Complementary Prompt (2026.eacl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have demonstrated strong performance across diverse tasks, but their performance varies significantly across different prompts. |
| Approach: | They propose a framework that reframes prompt engineering as a classification problem. |
| Outcome: | The proposed framework improves answer quality by up to 26.5% in win rate compared to prior methods while reducing latency by upto 1,956 times. |
Measuring and Mitigating Media Outlet Name Bias in Large Language Models (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing studies have explored the potential political biases of large language models, but limited attention has been devoted to the effects of media outlet names. |
| Approach: | They propose to quantify media outlet name biases in large language models and leverage this metric to develop an automated prompt optimization framework. |
| Outcome: | The proposed framework mitigates media outlet name biases, offering a scalable approach to enhancing the fairness of LLMs in news-related applications. |
Large Language Models are Students at Various Levels: Zero-shot Question Difficulty Estimation (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Recent advancements in educational platforms have emphasized the importance of personalized education. |
| Approach: | They propose a framework that utilizes large language models to represent students at various levels to estimate question difficulty with and without student question-solving records. |
| Outcome: | The proposed framework outperforms baseline models on the DBE-KT22 and ASSISTMents 2005–2006 benchmarks and shows a high correlation with the regressed IRT curve. |
Adaptive Compression of Word Embeddings (2020.acl-main)
Copied to clipboard
| Challenge: | Distributed representations of words have been an indispensable component for natural language processing (NLP) tasks. |
| Approach: | They propose a method that uses a code-book approach to represent words as discrete codes such as (8, 5, 2, 4). |
| Outcome: | The proposed method makes the highly compressed word embeddings without hurting the task accuracy. |
Representation Learning for Unseen Words by Bridging Subwords to Semantic Networks (2020.lrec-1)
Copied to clipboard
| Challenge: | Pre-trained word embeddings only include words that appeared in corpora where pre-tried embedds are learned. |
| Approach: | They propose a method to represent out-of-vocabulary words using subword information and knowledge. |
| Outcome: | The proposed method improves performance over baselines that only use subwords or knowledge to represent OOV words. |
Learning from Missing Relations: Contrastive Learning with Commonsense Knowledge Graphs for Commonsense Inference (2022.findings-acl)
Copied to clipboard
| Challenge: | Existing approaches to commonsense inference lack coverage and expressive diversity of commonsensense knowledge graphs. |
| Approach: | They propose a framework that contrasts sets of semantically similar and dissimilar events . they propose 'solar' framework that can be used to learn commonsense inference . |
| Outcome: | The proposed framework outperforms the state-of-the-art commonsense transformer on commonsensense inference by 1.84% on average among 8 metrics. |
Leveraging Knowledge Graph-Enhanced LLMs for Context-Aware Medical Consultation (2025.emnlp-main)
Copied to clipboard
| Challenge: | Recent advances in large language models have significantly influenced the field of online medical consultations, but critical challenges remain, such as the generation of hallucinated information and the integration of up-to-date medical knowledge. |
| Approach: | They propose a framework that combines retrieval-augmented generation with a structured medical knowledge graph. |
| Outcome: | The proposed framework outperforms baselines on two medical consultation datasets and shows significant improvements in hallucination reduction and clinical usefulness. |
KOAS: Korean Text Offensiveness Analysis System (2021.emnlp-demo)
Copied to clipboard
San-Hee Park, Kang-Min Kim, Seonhee Cho, Jun-Hyung Park, Hyuntae Park, Hyuna Kim, Seongwon Chung, SangKeun Lee
| Challenge: | morphological richness and complex syntax of Korean cause difficulties in neural model training. |
| Approach: | They propose a system that exploits contextual and linguistic features and estimates an offensiveness score for a Korean text. |
| Outcome: | The proposed system exploits both contextual and linguistic features and estimates an offensiveness score for a Korean text. |
Efficient Pre-training of Masked Language Model via Concept-based Curriculum Masking (2022.emnlp-main)
Copied to clipboard
| Challenge: | Masked language modeling (MLM) has been widely used for pre-training effective bidirectional representations but comes at a substantial training cost. |
| Approach: | They propose a concept-based curriculum masking method that evaluates the MLM difficulty of each token based on a carefully-designed linguistic difficulty criterion. |
| Outcome: | The proposed method significantly improves pre-training efficiency with the original BERT model at half the training cost. |
Multi-pretraining for Large-scale Text Classification (2020.findings-emnlp)
Copied to clipboard
| Challenge: | Existing methods for large-scale text classification involve excessive computation and memory overheads. |
| Approach: | They propose a self-supervised and weakly supervised pretraining frameworks for large-scale text classification with multiple categories. |
| Outcome: | The proposed framework improves on the self-supervised and weakly supervised methods while being computationally efficient. |
Learning to Generate Word Representations using Subword Information (C18-1)
Copied to clipboard
| Challenge: | Existing word-based approaches to learning word representations are blind to subword information in words. |
| Approach: | They propose a character-based word representation approach to learn word representations from characters. |
| Outcome: | The proposed model outperforms baseline models that regard words as atomic units . the proposed model achieves 18.5% improvement on average in perplexity for morphologically rich languages . |
Conflict and Overlap Classification in Construction Standards Using a Large Language Model (2025.naacl-industry)
Copied to clipboard
Seong-Jin Park, Youn-Gyu Jin, Hyun-Young Moon, Choi Bong-Hyuck, Lee Seung Hwan, Ohjoon Kwon, Kang-Min Kim
| Challenge: | Current manual approaches to analyzing overlapping or conflicting content are time-consuming, costly, and error-prone. |
| Approach: | They propose a large language model that uses a construction domain-adapted large language for the semantic comparison of sentences in construction standards. |
| Outcome: | The proposed framework achieves 97.9% accuracy and 0.907 macro F1-score in classifying sentences from Korean construction standards as overlapping, conflicting, or neutral. |