UrbanDS: A Graph-Guided LLM Multi-Agent System for Data-Intensive Urban Tasks
Quick Answer
UrbanDS is a graph-guided LLM multi-agent system designed for data-intensive urban tasks, outperforming existing agents in benchmarks.
Quick Take
It utilizes a unified dataset graph and multiple specialized agents to efficiently process urban data, demonstrating effectiveness in real-world applications in Wuhan's Dongxihu District.
Key Points
- UrbanDS constructs a unified dataset graph to organize dataset skills and relationships.
- The system includes specialized agents for profiling, planning, execution, and reporting.
- UrbanDS-Bench benchmarks its performance on urban data science tasks.
- Experiments show UrbanDS consistently outperforms existing data science agents.
- Successfully deployed in Wuhan's urban operations platform, showcasing real-world effectiveness.
DeepSignal Analysis
What happened
UrbanDS is a multi-agent system that employs a graph-guided approach to manage data-intensive urban tasks. It features a unified dataset graph and specialized agents for data profiling, relationship identification, planning, execution, and reporting. The system has been tested against benchmarks and deployed in Wuhan's Dongxihu District.
Key evidence
- UrbanDS utilizes a unified dataset graph to organize dataset skills and their relationships, facilitating efficient data processing.
- The system includes multiple agents, such as a Data Profiling Agent and a Planner Agent, which work together to handle urban data tasks.
- Experiments indicate that UrbanDS outperforms existing data science agents on both general and urban benchmarks, demonstrating its effectiveness in real-world applications.
Why it matters
The development of UrbanDS addresses significant challenges in urban data management, particularly the complexities of large-scale and heterogeneous datasets. By leveraging a graph-guided architecture, it enhances the ability to discover and utilize relevant information, which is crucial for effective urban planning and operations. Its deployment in a real-world setting further validates its practical utility and potential impact on urban data science.
Paper Resources
📖 Reader Mode
~2 min readAbstract:Large language model (LLM) agents have been widely applied in automating data science tasks. However, existing methods typically rely on a limited set of provided datasets, and they face challenges in data-intensive scenarios that require discovering and leveraging relevant information from large-scale and heterogeneous data repositories. Urban tasks are representative examples of such scenarios, as urban data are not only large-scale and multi-sourced, but also exhibit complex spatial, temporal, and semantic relationships. To address these challenges, we propose UrbanDS, a graph-guided LLM multi-agent system for data-intensive urban tasks. We first construct a unified dataset graph to organize reusable dataset skills and the relationships among datasets. Specifically, we develop a Data Profiling Agent that constructs a skill for each dataset. Moreover, a Relation Agent identifies relationships among datasets and integrates these relationships into the dataset graph. At runtime, a Planner Agent retrieves task-relevant datasets from the graph and generates execution plans. Multiple Execution Agents then perform data processing and analysis, while their execution progress and intermediate results are shared through a common memory. Finally, a Report Agent synthesizes the experimental logs into a report, which can be further refined based on user feedback. To systematically evaluate the capability of agents in handling data-intensive urban scenarios, we further construct UrbanDS-Bench, an urban data science benchmark covering representative data analysis and modeling tasks. Experiments on both general and urban benchmarks demonstrate that UrbanDS consistently outperforms existing data science agents on data-intensive tasks. Furthermore, UrbanDS has been deployed on the urban operations platform of Dongxihu District, Wuhan, demonstrating its effectiveness in real-world urban applications.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2607.26724 [cs.AI] |
| (or arXiv:2607.26724v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2607.26724 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Zhilun Zhou [view email]
[v1]
Wed, 29 Jul 2026 10:14:02 UTC (835 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.AI
See more →HOBA: Hierarchical On-Policy Bidding Agents for Adaptive Online Advertising
HOBA (Hierarchical On-policy Bidding Agents) is a novel hierarchical reinforcement learning framework that enhances online advertising bidding systems by improving adaptability and reducing hyperparameter tuning costs. It utilizes a for hyperparameter inference, a SARSA agent for expert model selection, and a dynamic expert pool for bid execution, achieving a +3.6% increase in target cost during large-scale deployment and outperforming state-of-the-art baselines on AuctionNet.