| 基于多Agent的高校网络舆情监测与分析系统 |
| |
| 引用本文： | 于卫红.基于多Agent的高校网络舆情监测与分析系统[J].现代情报,2017,37(10):53-57. |
| |
| 作者姓名： | 于卫红 |
| |
| 作者单位： | 大连海事大学交通运输管理学院, 辽宁 大连 116026 |
| |
| 基金项目： | 2015年度教育部人文社会科学研究规划基金项目"微信环境下基于大数据的高校舆情监管机制研究"（项目编号：15YJAZH102）。 |
| |
| 摘 要： | 目的/意义]为校园舆情信息管理员、校园舆情监管办公室、相关领导等提供舆情信息服务及决策参考。方法/过程]在Jade平台下分析、设计了基于多Agent的高校网络舆情监测与分析系统。详细阐述了信息采集Agent、信息预处理Agent、舆情分析Agent、舆情简报生成Agent的工作机制。构建了由物理层、语法层和语义层组成的多Agent间的通信模型。为了提高Agent的舆情分析能力与分析结果的可视化展示能力，通过Rserve接口对Jade平台与R语言进行集成，使得Jade Agent可以灵活调用R语言丰富的算法包。结果/结论]文中给出了"高校百度贴吧热点分析"的系统运行实例。运行结果表明基于多Agent的高校网络舆情监测与分析系统充分利用了Agent的自主性、智能性和交互性，使得系统更加具有灵活性和实时性；同时系统以大数据思维为指导，并集成了大数据分析利器R语言，使得舆情分析更加全面、客观。
|
| 关 键 词： | 高校网络舆情 舆情监测与分析 多Agent Jade平台 R语言 |
University Network Public Opinion Monitoring and Analysis System Based on Multi-Agent |
| |
| Authors: | Yu Weihong |
| |
| Institution: | Transportation and Management College, Dalian Maritime University, Dalian 116026, China |
| |
| Abstract: | Purpose/Significance] To provide public opinion information service and decision-making reference for campus public opinion information administrators, campus public opinion supervision office and related leaders.Method/Process] Under Jade platform, a network public opinion monitoring and analysis system based on multi Agent was analyzed and designed. The work mechanism of information collection Agent, information pre-processing Agent, public opinion a-nalysis Agent and public opinion report Agent were elaborated in detail. A multi Agent communication model consisting of physical layer, syntax layer and semantic layer was constructed. In order to improve Agents' public opinion analysis ability and the ability of analysis result visualization display, Jade platform was integrated with R language through Rserve inter-face, so that Jade Agent could flexibly call rich algorithms of R language.Result/Conclusion] As a system running in-stance, "hot spot analysis of university Baidu Post Bar" was given. The running results showed that University Network Public Opinion Monitoring and Analysis System Based on Multi-Agent takes full advantage of Agent good properties such as autonomy, intelligence and interactivity, so that the system became more flexible and real-time. Meanwhile, taking big data thinking as a guide and integrating R language made public opinion analysis more comprehensive and objective. |
| |
| Keywords: | university network public opinion public opinion monitoring and analysis Multi-Agent Jade plat-form R language |
|
| 点击此处可从《现代情报》浏览原始摘要信息 |
| 点击此处可从《现代情报》下载免费的PDF全文 |
