基于智能体标注的多模态政务服务热线数据集

一、基本信息
本数据集由江苏钟吾大数据发展集团自主打造，聚焦12345政务热线场景，构建文本+音频双模态数据体系，全链路能力达行业领先水平。依托自研标注智能体+人工审核，搭建专业标注体系，含6类诉求、五级分类等多维度标注，兼顾效率与准确性，可全面赋能AI大模型全建设阶段，为政务AI应用筑牢高质量数据底座。
二、覆盖范围
数据集覆盖宿城区全量12345政务数据。
三、典型应用场景
一是政务热线工单智能分拣，训练模型攻克分类难题，提升分拣效率与准确率，降低返工率、压缩响应时间；二是政务敏感事件智能预警，精准识别涉稳、公共安全等敏感工单并分级预警，提升处置效率、防范突发情况；三是政务诉求热点分析，挖掘数据构建分析与评估体系，为政务资源配置提供决策依据，推动解决民生热点问题；同时可延伸至城市治理精准决策、民生服务趋势研判等场景，为政务大模型训练提供核心数据保障。

I. Basic Information
This dataset is independently developed by Jiangsu Zhongwu Big Data Development Group, focusing on the 12345 government service hotline scenario, and builds a text+audio bimodal data system. Its full-link capabilities rank at the industry-leading level. Relying on self-developed annotation AI Agents and manual reviews, it has established a professional annotation system with multi-dimensional annotations including 6 types of citizen appeals and five-level classification, balancing efficiency and accuracy. It can fully empower the entire development cycle of AI Large Language Models (LLMs), and lay a high-quality data foundation for government AI applications.

II. Coverage Scope
This dataset covers all 12345 government service data in Suqian District.

III. Typical Application Scenarios
First, intelligent sorting of government hotline work orders: train models to tackle classification challenges, improve sorting efficiency and accuracy, reduce rework rates and shorten response time.
Second, intelligent early warning for government sensitive incidents: accurately identify sensitive work orders related to social stability, public safety and other aspects, and conduct graded early warning, so as to improve disposal efficiency and prevent emergencies.
Third, hotspot analysis of government appeals: mine data to build an analysis and evaluation system, provide decision-making basis for government resource allocation, and promote the resolution of people's livelihood hotspots.
In addition, it can be extended to scenarios such as precise decision-making for urban governance and trend analysis of people's livelihood services, providing core data support for the training of government LLMs.

搜集汇总

背景与挑战

背景概述

该数据集是一个多模态政务服务热线数据集，采用智能体标注技术构建。