# Synergizing Unsupervised Episode Detection with LLMs for Large-Scale News Events

arXiv: https://arxiv.org/abs/2408.04873 · HTML: https://mlsanity.com/paper/2408.04873/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 22 of its 45 references that are in the map.
- First posted on arXiv 9 Aug 2024.

## Builds on (most cited first, 22 of 22)

- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [Lost in the Middle: How Language Models Use Long Contexts](https://mlsanity.com/paper/2307.03172/) (2023)
- [LooGLE: Can Long-Context Language Models Understand Long Contexts?](https://mlsanity.com/paper/2311.04939/) (2023)
- [Event Extraction by Answering (Almost) Natural Questions](https://mlsanity.com/paper/2004.13625/) (2020)
- [Document-Level Event Argument Extraction by Conditional Generation](https://mlsanity.com/paper/2104.05919/) (2021)
- [Text2Event: Controllable Sequence-to-Structure Generation for End-to-end Event Extraction](https://mlsanity.com/paper/2106.09232/) (2021)
- [Jointly Multiple Events Extraction via Attention-based Graph Information Aggregation](https://mlsanity.com/paper/1809.09078/) (2018)
- [Joint Constrained Learning for Event-Event Relation Extraction](https://mlsanity.com/paper/2010.06727/) (2020)
- [X-Class: Text Classification with Extremely Weak Supervision](https://mlsanity.com/paper/2010.12794/) (2020)
- [GATE: Graph Attention Transformer Encoder for Cross-lingual Relation and Event Extraction](https://mlsanity.com/paper/2010.03009/) (2020)
- [Discriminative Topic Mining via Category-Name Guided Text Embedding](https://mlsanity.com/paper/1908.07162/) (2019)
- [Examining the State-of-the-Art in News Timeline Summarization](https://mlsanity.com/paper/2005.10107/) (2020)
- [Analogous Process Structure Induction for Sub-event Sequence Prediction](https://mlsanity.com/paper/2010.08525/) (2020)
- [Deep Structured Neural Network for Event Temporal Relation Extraction](https://mlsanity.com/paper/1909.10094/) (2019)
- [A Temporally Sensitive Submodularity Framework for Timeline Summarization](https://mlsanity.com/paper/1810.07949/) (2018)
- ["What Are You Trying to Do?" Semantic Typing of Event Processes](https://mlsanity.com/paper/2010.06724/) (2020)
- [Automatic Event Salience Identification](https://mlsanity.com/paper/1809.00647/) (2018)
- [Memory and Knowledge Augmented Language Models for Inferring Salience in Long-Form Stories](https://mlsanity.com/paper/2109.03754/) (2021)
- [Open-Vocabulary Argument Role Prediction for Event Extraction](https://mlsanity.com/paper/2211.01577/) (2022)
- [Hierarchical Quantized Representations for Script Generation](https://mlsanity.com/paper/1808.09542/) (2018)
- [Unsupervised Story Discovery from Continuous News Streams via Scalable Thematic Embedding](https://mlsanity.com/paper/2304.04099/) (2023)
- [MEGClass: Extremely Weakly Supervised Text Classification via Mutually-Enhancing Text Granularities](https://mlsanity.com/paper/2304.01969/) (2023)

## Built on by (most cited first, 1 of 1)

- [A Survey on Retrieval And Structuring Augmented Generation with Large Language Models](https://mlsanity.com/paper/2509.10697/) (2025)

## Abstract (arXiv)

State-of-the-art automatic event detection struggles with interpretability and adaptability to evolving large-scale key events -- unlike episodic structures, which excel in these areas. Often overlooked, episodes represent cohesive clusters of core entities performing actions at a specific time and location; a partially ordered sequence of episodes can represent a key event. This paper introduces a novel task, episode detection, which identifies episodes within a news corpus of key event articles. Detecting episodes poses unique challenges, as they lack explicit temporal or locational markers and cannot be merged using semantic similarity alone. While large language models (LLMs) can aid with these reasoning difficulties, they suffer with long contexts typical of news corpora. To address these challenges, we introduce EpiMine, an unsupervised framework that identifies a key event's candidate episodes by leveraging natural episodic partitions in articles, estimated through shifts in discriminative term combinations. These candidate episodes are more cohesive and representative of true episodes, synergizing with LLMs to better interpret and refine them into final episodes. We apply EpiMine to our three diverse, real-world event datasets annotated at the episode level, where it achieves a 59.2% average gain across all metrics compared to baselines.