大数据

• •    

基于时间编码LSTM的高校舆情热点趋势预测研究

易杰, 曹腾飞, 黄肖翰, 张子震   

  1. 青海大学计算机技术与应用系,青海省西宁市 810016
  • 作者简介:易杰(1998-),男,硕士研究生,现就读于青海大学计算机技术与应用系,主要研究方向为深度学习、数据挖掘。 曹腾飞(1987-),男(通信作者),博士,青海大学计算机技术与应用系副教授、硕导,CCF会员,主要研究方向为舆情分析、服务推荐与管理(caotf@qhu.edu.cn)。 黄肖翰(1999-),男,硕士研究生,现就读于青海大学计算机技术与应用系,主要研究方向为深度学习、数据挖掘。张子震(1994-),男,硕士研究生,现就读于青海大学计算机技术与应用系,主要研究方向为强化学习、舆情分析。

Research on trend prediction of time coded LSTM based public opinion hot spots in universities

YI Jie, CAO Tengfei, HUANG Xiaohan, ZHANG Zizhen   

  1. Department of Computer Technology and Applications,Qinghai University,Xining 810016,China

摘要: 随着互联网技术的发展,网络舆情热点信息能在短时间内得到迅速传播。预测舆情热点的发展趋势,有助于高校对于学生思想健康的分析管理,也是当下网络舆情信息研究领域的重要问题。本文针对微博中的舆情信息文本,构建基于时间编码的长短期记忆网络(Long Short Term Memory, LSTM)高校舆情热点趋势预测模型,并通过实验对比支持向量机(Support Vector Machines, SVM)、循环神经网络(Recurrent Neural Network, RNN)两种模型的预测效果,验证了加入时间编码参数后的LSTM算法在舆情趋势预测上的准确性。最后,利用微博中高校实时舆情事件对本文构建的模型预测效果进行评估,并动态调整评估参数,实现了对评估性能的优化,预测效果得到了显著提升。

关键词: LSTM, 热点预测, 高校舆情, 时序数据, 时间编码

Abstract: With the development of Internet technology, network public opinion hot information can be quickly spread in a short time. Predicting the development trend of public opinion hot spots is helpful to the analysis and management of college students' ideological health, and it is also an important issue in the field of network public opinion information research. Aiming at the public opinion information text in microblog, this paper constructs the hot trend prediction model of colleges and universities based on time coded Long Short Term Memory (LSTM) comparing with Support Vector Machines (SVM), Recurrent Neural Network (RNN) through experiments. The effectiveness of the two models verifies the superiority of time coded LSTM. Finally, the prediction effect of time coded LSTM is evaluated by using the real-time public opinion events of colleges and universities in microblog, and the evaluation parameters are dynamically adjusted to optimize the performance of the evaluation, and the prediction effect is improved significantly.

Key words: LSTM, hot spot prediction, public opinion of colleges and universities, time series data, time coded

No Suggested Reading articles found!