TimelineAgent builds event timelines from many documents by query
TimelineAgent: An LLM-based Framework for Query-Centric Event Timeline Extraction Across Multiple Documents
Artificial Intelligence
Summary
Understanding how events unfold over time from many documents is hard, especially when questions vary. The authors present TimelineAgent, which uses a large language model with multiple agents working in steps to gather and organize event details. They also created a special test set to measure how well different methods work on this kind of timeline task. TimelineAgent did better than previous approaches, showing promise for real-world uses where you want to track event sequences based on different queries.
What this means in practice
- •For intelligence analysts: Build detailed event timelines from diverse sources focused on specific investigative queries to support analysis.
- •For journalism teams: Create accurate, query-driven timelines of complex events spanning many news articles to improve reporting clarity.
Authors
Qingkai Min, Yunxin Bao, Yifeng Lu, Zhenyu Feng, Yulong Chen, Andreas Vlachos, Yue Zhang
Abstract
Query-centric cross-document event timeline extraction aims to construct event timelines from multiple documents under diverse query focuses. The task is very challenging due to the heterogeneous event modeling and reasoning requirements arising from both the intrinsic complexity of cross-document timeline construction and the diversity of query focuses. In this paper, we propose TimelineAgent, an LLM-based multi-agent framework featuring a multi-stage reasoning workflow, where heterogeneous event features are organized into reusable skills and orchestrated through a goal-oriented protocol. To facilitate evaluation, we construct a benchmark covering diverse query focuses. Experimental results show that TimelineAgent consistently outperforms existing methods, offering a practical framework for complex real-world timeline construction.