图书情报工作 ›› 2018, Vol. 62 ›› Issue (3): 71-77.DOI: 10.13266/j.issn.0252-3116.2018.03.009

• 情报研究 • 上一篇    下一篇

微博内向型传导热点发现与预测算法研究

王林森1, 王学义2   

  1. 1. 西南医科大学人文与管理学院 泸州 646000;
    2. 西南财经大学中国西部经济研究中心 成都 611130
  • 收稿日期:2017-09-07 修回日期:2017-11-15 出版日期:2018-02-05 发布日期:2018-02-05
  • 作者简介:王林森(ORCID:0000-0002-1203-9119),讲师,博士,E-mail:wanglinsen1978@126.com;王学义,教授,博士后,博士生导师。
  • 基金资助:
    本文系国家社会科学基金青年项目"群体性事件管理推演与应对措施研究"(项目编号:14CGL050)和中央高校基本科研业务费专项资金"基于复杂网络的微博用户关系发现研究"(项目编号:JBK141106)研究成果之一。

Research on the Detecting and Prognosis Algorithm of the Micro-blog Hotspot Transmission Inward Node

Wang Linsen1, Wang Xueyi2   

  1. 1. Humanlties and Management Sciences School of Southwest Medical University, Luzhou 646000;
    2. West China Center for Economic Research Southwestern University of Finance and Economics, Chengdu 611130
  • Received:2017-09-07 Revised:2017-11-15 Online:2018-02-05 Published:2018-02-05

摘要: [目的/意义]为应对微博内向型传导热点生灭速度快、热点特征不明显等问题,研究新型的微博内向型传导热点发现与预测算法。[方法/过程]针对上述问题,基于复杂网络分析方法,构建微博传导热点预测算法,该算法通过复杂网络节点模型扩展生成微博传导节点模型,发现内向型传导节点的传导子网;通过对传导节点序列实施热功率计算,对其信息传导覆盖范围以及未来影响力进行预测,进行传导热点发现及预测。[结果/结论]数据实验表明,该算法较之目前常用的热点预测算法,具有较高的传导热点覆盖率和准确率,且耗时较少,性价比较高。

关键词: 舆情监测, 微博, 复杂网络, 传导节点, 热点预测

Abstract: [Purpose/significance]In order to deal with the quick birth-death process and unconspicuous hotspot characteristics of information transmission nodes, a new detecting and prognosis system is proposed based on complex networks. [Method/process]In order to deal with those problems, a novel micro-blog information transmission hotspot inward node prognosis algorithm was proposed as the IPIN algorithm based on complex network analysis methods. This paper used this algorithm to build a model with complex network node relations, and found relation sub-networks by related nodes. Then, the thermal power spectrum computing was used to dope out information transmission ranges and prospective effects. [Result/conclusion]Data experiment results prove that the IPIN algorithm has higher hot transmission node coverage rate, accuracy rate and better cost-performance than those of the SNMA algorithm.

Key words: public opinion monitoring, micro-blog, complex network, information transmission node, hotspot prognosis

中图分类号: