MUDY: Multi-Granular Dynamic Candidate Contextualization for Unsupervised Keyphrase Extraction
sigir26-158 · Full research paper · Hyeongu Kang, Susik Yoon
Language Model Pre-training and Adap · 100%
Abstract
Keyphrase extraction aims to automatically identify concise phrases that effectively represent the content of a document. While recent methods leveraging pre-trained language models (PLMs) have significantly improved the extraction of keyphrases with strong global semantic relevance, they often fall short in capturing the local contextual importance of keyphrases tied to specific subtopics dispersed in a document. In this paper, we propose a novel context-centric framework, MUDY, that effectively captures multi-granular contextual salience of candidate keyphrases. MUDY employs two complementary components: (1) a prompt-based scoring that estimates the generation likelihood of each candidate keyphrase, augmented with candidate-aware weighting to better reflect its local contextual importance, and (2) a self-attention-based scoring that utilizes multi-granular attention patterns from PLMs to assess candidate significance at both the document-wide and segment-specific levels. Evaluations on four real-world datasets demonstrate that MUDY outperforms state-of-the-art baselines in top-k accuracy at various cutoff thresholds. In-depth quantitative and qualitative analyses further highlight the efficacy of context-centric keyphrase extraction with multi-granular saliency. For reproducibility, the source code of MUDY is available at https://github.com/HgKang1/MUDY.
Assigned reviewers
| # | Reviewer | Match | Why (closest evidence publications in the taxonomy) | |
|---|---|---|---|---|
| 1 | Weipeng Chen | 62 | 100 DataSculpt: A Holistic Data Management Framework for Long-Context LLMs Training (2025) Language Model Pre-training and Adaptation 100 Exploring Context Window of Large Language Models via Decomposed Positional Vectors (2024) Language Model Pre-training and Adaptation 100 KV Shifting Attention Enhances Language Modeling (2025) Language Model Pre-training and Adaptation |
|
| 2 | Di Wu 0054 | 57 | 100 On Leveraging Encoder-only Pre-trained Language Models for Effective Keyphrase Generation (2024) Language Model Pre-training and Adaptation 100 Rethinking Model Selection and Decoding for Keyphrase Generation with Pre-trained Sequence-to-Sequence Models (2023) Language Model Pre-training and Adaptation 100 Active Instruction Tuning: Improving Cross-Task Generalization by Training on Prompt Sensitive Tasks (2023) Language Model Pre-training and Adaptation |
|
| 3 | Xi Ye 0003 | 56 | 100 Assessing Out-of-Domain Language Model Performance from Few Examples (2023) Language Model Pre-training and Adaptation 100 Explanation Selection Using Unlabeled Data for Chain-of-Thought Prompting (2023) Language Model Pre-training and Adaptation 100 Can Explanations Be Useful for Calibrating Black Box Models? (2022) Language Model Pre-training and Adaptation |
Area chair
| Area chair | Match | |
|---|---|---|
| Xiaoye Qu | 49 |
Area-chair candidates (3)
| Xiaoye Qu assigned | 49 | 15 papers | |
| Yi R. Fung 0001 | 39 | 10 papers | |
| Xiaoyi Zeng | 33 | 11 papers |
Candidates from the panel ranked by match score
| # | Reviewer | Match | Load | Why | |
|---|---|---|---|---|---|
| 1 | Weipeng Chen assigned | 62 | 3 | 100 DataSculpt: A Holistic Data Management Framework for Long-Context LLMs Training (2025) Language Model Pre-training and Adaptation 100 Exploring Context Window of Large Language Models via Decomposed Positional Vectors (2024) Language Model Pre-training and Adaptation |
|
| 2 | Di Wu 0054 assigned | 57 | 3 | 100 On Leveraging Encoder-only Pre-trained Language Models for Effective Keyphrase Generation (2024) Language Model Pre-training and Adaptation 100 Rethinking Model Selection and Decoding for Keyphrase Generation with Pre-trained Sequence-to-Sequence Models (2023) Language Model Pre-training and Adaptation |
|
| 3 | Xi Ye 0003 assigned | 56 | 3 | 100 Assessing Out-of-Domain Language Model Performance from Few Examples (2023) Language Model Pre-training and Adaptation 100 Explanation Selection Using Unlabeled Data for Chain-of-Thought Prompting (2023) Language Model Pre-training and Adaptation |
|
| 5 | Huiqiang Jiang | 45 | 3 | 100 Decomposed Meta-Learning for Few-Shot Sequence Labeling (2024) Language Model Pre-training and Adaptation 100 Chain-of-Model Learning for Language Model (2025) Language Model Pre-training and Adaptation |
|
| 6 | Ori Yoran | 44 | 3 | 100 Evaluating the Ripple Effects of Knowledge Editing in Language Models (2024) Language Model Pre-training and Adaptation 100 LMEnt: A Suite for Analyzing Knowledge in Language Models from Pretraining Data to Representations (2026) Language Model Pre-training and Adaptation |
|
| 7 | Yanzhao Zhang | 42 | 3 | 100 Unsupervised Boundary-Aware Language Model Pretraining for Chinese Sequence Labeling (2022) Language Model Pre-training and Adaptation 100 Chinese Sequence Labeling with Semi-Supervised Boundary-Aware Language Model Pre-training (2024) Language Model Pre-training and Adaptation |
|
| 8 | Weiwei Sun 0001 | 41 | 3 | 100 How Large Language Models Encode Context Knowledge? A Layer-Wise Probing Study (2024) Language Model Pre-training and Adaptation 100 Improving the Robustness of Large Language Models via Consistency Alignment (2024) Language Model Pre-training and Adaptation |
|
| 9 | Archiki Prasad | 40 | 3 | 100 GrIPS: Gradient-free, Edit-based Instruction Search for Prompting Large Language Models (2023) Language Model Pre-training and Adaptation 100 Multi-Attribute Steering of Language Models via Targeted Intervention (2025) Language Model Pre-training and Adaptation |
|
| 11 | Haiyang Xu 0001 | 38 | 3 | 100 Exploiting Presentative Feature Distributions for Parameter-Efficient Continual Learning of Large Language Models (2025) Language Model Pre-training and Adaptation 43 TinyChart: Efficient Chart Understanding with Program-of-Thoughts Learning and Visual Token Merging (2024) Multimodal Evaluation and Benchmarks |
|
| 12 | Dan Luo 0004 | 38 | 3 | 100 Mitigating Language Confusion through Inference-time Intervention (2025) Language Model Pre-training and Adaptation 62 From Anchors to Answers: A Novel Node Tokenizer for Integrating Graph Structure into Large Language Models (2025) KG-Augmented LLM Reasoning |
|
| 13 | Kaixin Ma | 38 | 3 | 100 Exploring Strategies for Generalizable Commonsense Reasoning with Pre-trained Models (2021) Language Model Pre-training and Adaptation 100 Transferring Procedural Knowledge Across Commonsense Tasks (2023) Language Model Pre-training and Adaptation |
|
| 14 | Venktesh V | 37 | 3 | 100 Topic Aware Contextualized Embeddings for High Quality Phrase Extraction (2022) Language Model Pre-training and Adaptation 100 Sample Efficient Demonstration Selection for In-Context Learning (2025) Language Model Pre-training and Adaptation |
|
| 15 | Xinying Qian | 37 | 3 | 100 Grow-on-Demand: Sparse and Adaptive Expert Expansion for Continual Instruction Tuning (2026) Language Model Pre-training and Adaptation 38 SMIR: Span-based multi-grained information refinement for joint multimodal entity-relation extraction (2026) Universal Multimodal Embedding and MLLM Retrieval |
|
| 16 | Junjie Zhang 0009 | 36 | 3 | 100 A Survey of Large Language Models (2026) Language Model Pre-training and Adaptation 63 A Pre-trained Plug-in Mixture-of-LoRAs Model for Transferable Sequential Recommendation (2025) Language Model Pre-training and Adaptation |
|
| 17 | Haokun Wen | 36 | 3 | 38 FiRE: Enhancing MLLMs with Fine-Grained Context Learning for Complex Image Retrieval (2025) Universal Multimodal Embedding and MLLM Retrieval 38 Fine-grained Textual Inversion Network for Zero-Shot Composed Image Retrieval (2024) Universal Multimodal Embedding and MLLM Retrieval |
|
| 18 | Zihao Yue | 35 | 3 | 43 Learning Descriptive Image Captioning via Semipermeable Maximum Likelihood Estimation (2023) Multimodal Evaluation and Benchmarks 43 ChartM3: Benchmarking Chart Editing with Multimodal Instructions (2025) Multimodal Evaluation and Benchmarks |
|
| 19 | Jinheon Baek | 35 | 3 | 100 System Prompt Optimization with Meta-Learning (2025) Language Model Pre-training and Adaptation 62 KALA: Knowledge-Augmented Language Model Adaptation (2022) KG-Augmented LLM Reasoning |
|
| 20 | Farima Fatahi Bayat | 35 | 3 | 100 CompactIE: Compact Facts in Open Information Extraction (2022) Language Model Pre-training and Adaptation 23 From Proof to Program: Characterizing Tool-Induced Reasoning Hallucinations in Large Language Models (2026) Reward Modeling and Process Supervision |
|
| 21 | Adam Jatowt | 35 | 3 | 100 BiTimeBERT: Extending Pre-Trained Language Representations with Bi-Temporal Information (2023) Language Model Pre-training and Adaptation 100 Towards Effective Time-Aware Language Representation: Exploring Enhanced Temporal Understanding in Language Models (2025) Language Model Pre-training and Adaptation |
|
| 22 | Shuaiqiang Wang | 34 | 3 | 66 S2phere: Semi-Supervised Pre-training for Web Search over Heterogeneous Learning to Rank Data (2023) Language Model Pre-training and Adaptation 66 RankElectra: Semi-supervised Pre-training of Learning-to-Rank Electra for Web-scale Search (2025) Language Model Pre-training and Adaptation |
|
| 23 | Lichan Hong | 34 | 3 | 66 Improving Data Efficiency for Recommenders and LLMs (2024) Language Model Pre-training and Adaptation 36 Better Generalization with Semantic IDs: A Case Study in Ranking for Recommendations (2024) Item Tokenization and Semantic IDs |
|
| 24 | Minghui Fang 0002 | 33 | 3 | 38 Multimodal Conditional Retrieval with High Controllability (2025) Universal Multimodal Embedding and MLLM Retrieval 38 CART: A Generative Cross-Modal Retrieval Framework With Coarse-To-Fine Semantic Modeling (2025) Universal Multimodal Embedding and MLLM Retrieval |
|
| 25 | Hao Liu 0078 | 33 | 3 | 38 Mitigating Fine-tuning Bias: A Parameter-Efficient Debiasing Framework for Large Language Models (2026) Bias Mitigation Methods 37 DA-DAN: A Dual Adversarial Domain Adaption Network for Unsupervised Non-overlapping Cross-domain Recommendation (2024) Domain Adaptation and Disentangled Representation for CDR |
|
| 26 | Yang Zhang 0072 | 33 | 3 | 43 Latent Inter-User Difference Modeling for LLM Personalization (2025) Personalized LLM Adaptation 39 Don't Start Over: A Cost-Effective Framework for Migrating Personalized Prompts Between LLMs (2026) Efficiency and Deployment |
|
| 27 | Ed H. Chi | 33 | 3 | 66 Improving Data Efficiency for Recommenders and LLMs (2024) Language Model Pre-training and Adaptation 36 Better Generalization with Semantic IDs: A Case Study in Ranking for Recommendations (2024) Item Tokenization and Semantic IDs |
|
| 29 | Corby Rosset | 33 | 3 | 67 Axiomatic Preference Modeling for Longform Question Answering (2023) Retrieval-Augmented and Open-Domain QA 65 Dodo: Dynamic Contextual Compression for Decoder-only LMs (2024) Learned Sparse Retrieval Efficiency |
|
| 30 | Qisheng Hu | 33 | 3 | 100 STARE at the Structure: Steering ICL Exemplar Selection with Structural Alignment (2025) Language Model Pre-training and Adaptation 25 Just What You Desire: Constrained Timeline Summarization with Self-Reflection for Enhanced Relevance (2025) Multi-hop and Complex Reasoning QA |
|
| 31 | Xinyu Ma 0001 | 33 | 3 | 37 B-PROP: Bootstrapped Pre-training with Representative Words Prediction for Ad-hoc Retrieval (2021) Sparse and Dense Retrieval Encoders 37 PROP: Pre-training with Representative Words Prediction for Ad-hoc Retrieval (2021) Sparse and Dense Retrieval Encoders |
|
| 32 | Ruqing Zhang 0001 | 33 | 3 | 37 B-PROP: Bootstrapped Pre-training with Representative Words Prediction for Ad-hoc Retrieval (2021) Sparse and Dense Retrieval Encoders 37 PROP: Pre-training with Representative Words Prediction for Ad-hoc Retrieval (2021) Sparse and Dense Retrieval Encoders |
|
| 33 | Shengyao Zhuang | 33 | 3 | 43 A Setwise Approach for Effective and Highly Efficient Zero-shot Ranking with Large Language Models (2024) LLM-Based Reranking 43 Where Relevance Emerges: A Layer-Wise Study of Internal Attention for Zero-Shot Re-Ranking (2026) LLM-Based Reranking |