Papers by Jiyoung Lee
Journalism-Guided Agentic In-context Learning for News Stance Detection (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing stance detection research on news content is limited to short texts and high-resource languages. |
| Approach: | They propose a dataset for article-level stance detection that integrates viewpoints into recommendation algorithms and a framework that employs a language model agent to predict the stances of key structural segments. |
| Outcome: | The proposed framework outperforms existing methods in identifying article stances and uncovering patterns of media bias. |
Single Ground Truth Is Not Enough: Adding Flexibility to Aspect-Based Sentiment Analysis Evaluation (2025.naacl-long)
Copied to clipboard
| Challenge: | Aspect-based sentiment analysis (ABSA) is a challenging task of extracting sentiments along with their corresponding aspects and opinion terms from text. |
| Approach: | They propose a pipeline that expands existing evaluation sets by adding alternative valid terms for aspect and opinion. |
| Outcome: | The proposed evaluation set uncovers the capabilities of large language models (LLMs) in ABSA tasks, which is concealed by the single-answer GT sets. |
The Fallacy of Echo Chambers: Analyzing the Political Slants of User-Generated News Comments in Korean Media (D19-55)
Copied to clipboard
| Challenge: | a new study analyzes the political slants of user comments on partisan media in Korea . the classifiers detect political leaning on conservative and liberal news outlets . |
| Approach: | They built a BERT-based classifier to detect political leaning of short comments . they found a high presence of conservative bias on conservative and liberal news outlets . |
| Outcome: | The proposed classifier produced an F1 score of 0.83 for 21.6K comments . it shows that more liberals comment on stories resonating with their political perspectives . |
Specializing Multi-domain NMT via Penalizing Low Mutual Information (2022.emnlp-main)
Copied to clipboard
| Challenge: | Multi-domain Neural Machine Translation (NMT) trains a single model with multiple domains. |
| Approach: | They propose a method that penalizes low MI to be higher for domain-specific NMTs. |
| Outcome: | The proposed method achieves state-of-the-art performance among current models . it also promotes low MI to be higher resulting in domain-specialized multi-domain NMT. |
KorNAT: LLM Alignment Benchmark for Korean Social Values and Common Knowledge (2024.findings-acl)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) must possess an understanding of the nation’s culture and basic knowledge. |
| Approach: | They propose to construct a national alignment benchmark, KorNAT, which measures the alignment between an LLM and a targeted country from two perspectives: social value alignment and common knowledge alignment. |
| Outcome: | The proposed model passes the national alignment score of 7 LLMs, indicating there is room for improvement. |
K/DA: Automated Data Generation Pipeline for Detoxifying Implicitly Offensive Language in Korean (2025.acl-long)
Copied to clipboard
| Challenge: | Language detoxification involves removing toxicity from offensive language. |
| Approach: | They propose an automated pipeline to generate offensive language with implicit offensiveness and trend-aligned slang. |
| Outcome: | The proposed dataset exhibits high pair consistency and greater implicit offensiveness compared to existing Korean datasets and demonstrates applicability to other languages. |