Papers by Yo-Sub Han
Detecting context abusiveness using hierarchical deep learning (D19-50)
Copied to clipboard
| Challenge: | Abusive text is a serious problem in social media and causes many issues among users . a model that detects text abusiveness in context without explicit abusive words is challenging . |
| Approach: | They propose to use an abusive lexicon to determine the existence of an abusive word in text . they combine local and global features to evaluate the model using benchmark data . |
| Outcome: | The proposed model outperforms all previous models for detecting abusiveness in text without abusive words. |
DITTO: A Spoofing Attack Framework on Watermarked LLMs via Knowledge Distillation (2026.eacl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) generate coherent, human-like text at scale, but raises concerns about authenticity and trust. |
| Approach: | They propose a threat of watermark spoofing that allows a malicious model to generate text containing the authentic-looking watermark of a trusted, victim model. |
| Outcome: | The proposed attack repurposes watermark radioactivity from a discoverable trait into an attack vector and replicates it. |
Self-Training using Rules of Grammar for Few-Shot NLU (2021.findings-emnlp)
Copied to clipboard
| Challenge: | Existing methods for learning natural language understanding are limited in low-resource settings. |
| Approach: | They propose to use rules of grammar to construct and expand rules of grammatical structure of data without human involvement. |
| Outcome: | The proposed approach outperforms state-of-the-art methods in three benchmark datasets. |
A Regex Minimization Benchmark: A PSPACE-Complete Challenge for Language Models (2026.eacl-long)
Copied to clipboard
| Challenge: | Language models (LMs) have demonstrated impressive reasoning capabilities across domains . but their ability to handle PSPACE-complete problems remains underexplored . a new benchmark for regex minimization is proposed to evaluate LMs' reasoning capabilities . |
| Approach: | They propose a benchmark for regex minimization to evaluate LMs' reasoning power . they use a million regexes paired with their minimal equivalents to evaluate their performance . |
| Outcome: | The proposed model can solve NP-complete problems, but their ability to handle PSPACE-complete ones remains underexplored. |
KatFishNet: Detecting LLM-Generated Korean Text through Linguistic Feature Analysis (2025.acl-long)
Copied to clipboard
| Challenge: | Detecting LLM-generated text is crucial for academic integrity, preventing plagiarism, protecting copyrights, ethical research practices. |
| Approach: | They propose a method specifically designed for Korean language to detect LLM-generated text . they examine spacing patterns, part-of-speech diversity, and comma usage . |
| Outcome: | The proposed method achieves an average of 19.78% higher AUC-ROC compared to the best-performing detection method. |
SuperST: Superficial Self-Training for Few-Shot Text Classification (2024.lrec-main)
Copied to clipboard
| Challenge: | In few-shot text classification, self-training relies on pseudo-labels to expand data, which has shown success, but can accumulate errors due to noisy pseudo-labeled data. |
| Approach: | They propose a method to mitigate noise in noisy pseudo-labeled data by applying superficial learning to noisy data and fine-tuning to less noisy data. |
| Outcome: | The proposed framework improves the classifier accuracy for few-shot text classification by 18.5% at most and 8% in average, compared with the state-of-the-art SSL baselines. |
Generalizable Implicit Hate Speech Detection Using Contrastive Learning (2022.coling-1)
Copied to clipboard
| Challenge: | Hate speech detection is challenging when there are insufficient lexical cues. |
| Approach: | They propose a contrastive learning method that pulls an implication and its corresponding posts close in representation space. |
| Outcome: | The proposed method improves on BERT and HateBERT benchmarks on three implicit hate speech benchmarks. |
Repairing Regex Vulnerabilities via Localization-Guided Instructions (2026.eacl-long)
Copied to clipboard
| Challenge: | Current approaches to regular expression denial of service (ReDoS) are hampered by a trade-off. |
| Approach: | They propose a framework to harness LLM generalization while enforcing reliability by localizing the vulnerable subpattern and generating a semanticallyequivalent fix for this isolated segment. |
| Outcome: | The proposed framework improves repair rates by 15.4%p over the state-of-the-art. |
GDA: Grammar-based Data Augmentation for Text Classification using Slot Information (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Recent studies suggest data augmentation approaches to resolve the low-resource problem in natural language processing tasks. |
| Approach: | They propose to use slot information to augment sentences using a set of injective relations between a sentence’s semantics and its syntactical structure to augment the dataset. |
| Outcome: | The proposed approach outperforms all other data augmentation methods by 19.38%. |
A Linguistics-Aware LLM Watermarking via Syntactic Predictability (2026.acl-long)
Copied to clipboard
| Challenge: | a central challenge remains balancing text quality against detection robustness. |
| Approach: | They propose a framework that aligns watermark strength with linguistic degrees of freedom . they use part-of-speech models to weaken the signal in grammatically constrained contexts . |
| Outcome: | The proposed framework outperforms existing methods in linguistic indeterminacy tests on languages . it weakens the watermark strength in grammatically constrained contexts and strengthens it in contexts with greater linguistic flexibility. |
Boosting Code Summarization by Embedding Code Structures (2022.coling-1)
Copied to clipboard
| Challenge: | Recent work on code summarization relies on structural information from the abstract syntax tree (AST) of source codes. |
| Approach: | They propose a program dependency graph (PDG) that represents the structure of a code more effectively. |
| Outcome: | The proposed model improves the performance of an out-of-domain benchmark dataset and the measure SBERT score. |
Online Infix Probability Computation for Probabilistic Finite Automata (P19-1)
Copied to clipboard
| Challenge: | Probabilistic finite automata (PFAs) are statistical language models used in natural language processing. |
| Approach: | They develop an asymptotic algorithm to compute the infix probabilities of each prefix of a string from streaming data. |
| Outcome: | The proposed algorithm improves the infix probabilities of a weighted automata from streaming data. |
TrapDoc: Deceiving LLM Users by Injecting Imperceptible Phantom Tokens into Documents (2025.findings-emnlp)
Copied to clipboard
| Challenge: | a growing number of LLMs have been used to provide reasoning, writing, text-editing capabilities. |
| Approach: | They propose a method to inject imperceptible phantom tokens into LLMs to deceive users . the technique generates outputs that appear plausible to users but are in fact incorrect . |
| Outcome: | a new method injects imperceptible tokens into documents to deceive users . the proposed framework is compared to baselines to show its effectiveness . |
Contrastive Learning with Keyword-based Data Augmentation for Code Search and Code Question Answering (2023.eacl-main)
Copied to clipboard
| Challenge: | Recent work on code search proposes data augmentation of queries for contrastive learning. |
| Approach: | They propose to augment query-code pairs with key words to preserve key words . they use keyDAC to fine-tune various pre-trained language models . |
| Outcome: | The proposed approach outperforms the current state-of-the-art in code search and question answering tasks. |
RV-HATE: Reinforced Multi-Module Voting for Implicit Hate Speech Detection (2026.acl-long)
Copied to clipboard
| Challenge: | a new framework for hate speech detection addresses implicit hate speech by tailoring the detection process to dataset-specific attributes. |
| Approach: | They propose a framework to account for the dataset-specific characteristics of hate speech datasets. |
| Outcome: | The proposed framework improves detection accuracy and provides interpretable insights into the distinctive features of each dataset. |
SoftRegex: Generating Regex from Natural Language Descriptions using Softened Regex Equivalence (D19-1)
Copied to clipboard
| Challenge: | Existing models that generate semantically correct regular expressions from NLs are not yet fully understood. |
| Approach: | They propose a model that rewards reinforcement learning based on the semantic equivalence between two regular expressions. |
| Outcome: | The proposed model reduces training time and produces state-of-the-art results on three benchmark datasets. |
ContractEval: A Benchmark for Evaluating Contract-Satisfying Assertions in Code Generation (2026.findings-acl)
Copied to clipboard
| Challenge: | Current evaluations measure functional correctness on well-formed inputs, but they filter out inputs that violate them. |
| Approach: | They propose a benchmark to evaluate whether generated code enforces preconditions . they use a neuro-symbolic pipeline to evaluate code with test cases . |
| Outcome: | The proposed benchmark aims to evaluate whether generated code enforces preconditions . it aims at achieving pass@k scores while ignoring those that violate them . |
Mondrian: A Framework for Logical Abstract (Re)Structuring (2025.emnlp-main)
Copied to clipboard
| Challenge: | Using the ABT structure, academic abstracts are structured to provide clear and concise prose, but a lack of clarity and logical coherence is a challenge for authors struggling with English proficiency or academic writing conventions. |
| Approach: | They propose a framework that identifies the key components of an abstract and reorients itself to properly reflect the ABT logical progression. |
| Outcome: | The proposed framework improves comprehensibility of academic writing, particularly for non-native English speakers, and is based on a human evaluation and automated metrics. |
AmpleHate: Amplifying the Attention for Versatile Implicit Hate Detection (2025.emnlp-main)
Copied to clipboard
| Challenge: | Current approaches to detect hate speech rely on contrastive learning to distinguish hate from non-hate sentences. |
| Approach: | They propose a novel approach to detect implicit hate speech by identifying explicit targets . they use a pretrained Named Entity Recognition model to capture explicit target information . |
| Outcome: | The proposed approach outperforms current methods and achieves faster convergence. |
ConPrompt: Pre-training a Language Model with Machine-Generated Data for Implicit Hate Speech Detection (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Existing pre-trained language models for hate speech detection are not specialized in implicit hate speech. |
| Approach: | They propose a pre-trained language model for implicit hate speech detection that leverages machine-generated data to train the model. |
| Outcome: | The proposed model can be trained on a massive hate speech dataset with positive samples . it can be generalized and reduce identity term bias, the authors show . |
MultiFix: Learning to Repair Multiple Errors by Optimal Alignment Learning (2021.findings-emnlp)
Copied to clipboard
| Challenge: | Existing approaches fix a single error in a line, but it is inevitable to iterate until no errors remain. |
| Approach: | They propose a sequence-to-sequence learning framework for fixing multiple program errors at once . they pare an erroneous program with an optimal alignment to the correct program . |
| Outcome: | The proposed approach achieves state-of-the-art on a dataset of 6,975 erroneous C programs . the proposed framework is based on an edit-distance-based data labeling approach . |
TCProF:Time-Complexity Prediction SSL Framework (2025.naacl-long)
Copied to clipboard
| Challenge: | determining the precise time complexity of a code is theoretically undecidable . determining time complexity is a challenging task in programming efficiency analysis . |
| Approach: | They propose a time-complexity prediction SSL framework that uses code snippets labeled with their time complexity classes to predict code time. |
| Outcome: | The proposed approach improves performance by 60% over self-training methods. |
Incremental Computation of Infix Probabilities for Probabilistic Finite Automata (D18-1)
Copied to clipboard
| Challenge: | Probabilistic finite automata are used to model distributions in natural language processing . a method that computes infix probabilities incrementally is proposed . |
| Approach: | They propose a method that computes infix probabilities incrementally for probabilistic finite automata . they propose to represent all the probabilities of matching strings as a series of transition matrix calculations . |
| Outcome: | The proposed method is theoretically faster than the previous method and better in practice. |
CodeComplex: Dataset for Worst-Case Time Complexity Prediction (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Reasoning ability of large language models (LLMs) is crucial in complex decision-making tasks. |
| Approach: | They propose to use code time complexity prediction to assess LLMs' reasoning ability. |
| Outcome: | The proposed dataset comprises 4,900 Java codes and an equivalent number of Python codes. |
Marking Code Without Breaking It: Code Watermarking for Detecting LLM-Generated Code (2026.findings-eacl)
Copied to clipboard
| Challenge: | Existing methods for identifying LLM-generated code are limited by syntax-critical tokens, which can introduce syntax errors. |
| Approach: | They propose a syntax-aware watermarking method that embeds watermarks only in non-syntactic tokens and preserves code integrity. |
| Outcome: | The proposed method outperforms baseline methods on Python, C++, and Java. |
Analyzing Offensive Language Dataset Insights from Training Dynamics and Human Agreement Level (2025.coling-main)
Copied to clipboard
| Challenge: | Existing models struggle to generalize themselves when applied to implicit cases, especially in out-of-domain scenarios. |
| Approach: | They propose a data refinement strategy that integrates human annotation agreement with model training dynamics to enhance model performance and generalization. |
| Outcome: | The proposed approach improves performance across multiple datasets and models while retaining ambiguous cases crucial for out-of-domain generalization. |
Query4Regex: Verifiable Regex Transformation through Formal Operations from NL and DSL Queries (2026.findings-eacl)
Copied to clipboard
| Challenge: | Large language models (LLMs) generate structured data, but their ability to precisely manipulate it remains relatively under-explored. |
| Approach: | They propose a benchmark to evaluate verifiable transformations on regexes . they use natural language instructions and a program-like domain-specific language that specifies the sequence of operations to evaluate LLMs. |
| Outcome: | The proposed benchmark compares LLM performance on natural language and DSL queries for regex manipulation. |
SharedCon: Implicit Hate Speech Detection using Shared Semantics (2024.findings-acl)
Copied to clipboard
| Challenge: | Recent studies suggest that classifying hateful posts in a binary manner may not address nuanced task of detecting implicit hate speech. |
| Approach: | They propose a contrastive learning approach that leverages shared semantics among data to detect implicit hate speech. |
| Outcome: | The proposed approach is based on a clustering-based contrastive learning approach with human-written implications or machine-generated augmented data. |
ATHENA: Mathematical Reasoning with Thought Expansion (2023.emnlp-main)
Copied to clipboard
| Challenge: | a mathematical word problem requires a method that articulates the problem, the lens through which models view human lexical expressions. |
| Approach: | They propose a method to mimic human thought expansion mechanisms to generate thought expansions . they show ATHENA can be used to generate reasonable thoughts by selecting valid pathways . |
| Outcome: | The proposed model is compelling in variant questions even when the informativeness in training examples is restricted. |