基于LSTM模型的食品安全网络舆情预警研究
|
摘要
近几年,食品安全网络舆情事件数量激增,引起了国家的高度重视。目前的食品安全网络舆情预警指标体系对主题属性和传播扩散指标考虑不全面,未深入考虑舆情自身特性和演化规律,而且目前的网络舆情预警模型也不能很好地考虑舆情不同特征之间的相互联系,导致舆情预警准确率不高。针对以上问题,提出包括主题属性、传播扩散等5个维度的指标体系,并在此基础上提出长短时记忆网络Re-LSTM模型,使用正则化方法约束网络中各单元输入权重并用softsign函数替代tanh激活函数。与其他经典模型对比,所构建的模型不仅能够提高预警准确率,而且还能够更好地避免梯度消失和过拟合问题。
In recent years, increasing events of internet public opinion on food security have attracted great attention of the Chinese government. The current early warning index system of internet public opinion on food security lacks a comprehensive consideration of theme attributes, propagation and diffusion indexes, and does not consider the inherent characteristics and evolution law of the public opinion in depth. Moreover, as the current early warning model of internet public opinion fails to take the interrelationship between different characteristics of the public opinion into consideration, which leads to the low accuracy of early warning of public opinion. Aiming at the above problems, we construct an index system consisting of five dimensions, including theme attributes and propagation and diffusion indexes. Based on this, we propose a regularization long short term memory(Re-LSTM) model, which uses the regularization method to constrain the input weight of each unit in the network, and replaces the tanh activation function by the softsign activation function. Compared with other classic models, the proposed model can not only improve the accuracy of early warning, but also better avoid the problem of gradient disappearance and overfitting.
In recent years, increasing events of internet public opinion on food security have attracted great attention of the Chinese government. The current early warning index system of internet public opinion on food security lacks a comprehensive consideration of theme attributes, propagation and diffusion indexes, and does not consider the inherent characteristics and evolution law of the public opinion in depth. Moreover, as the current early warning model of internet public opinion fails to take the interrelationship between different characteristics of the public opinion into consideration, which leads to the low accuracy of early warning of public opinion. Aiming at the above problems, we construct an index system consisting of five dimensions, including theme attributes and propagation and diffusion indexes. Based on this, we propose a regularization long short term memory(Re-LSTM) model, which uses the regularization method to constrain the input weight of each unit in the network, and replaces the tanh activation function by the softsign activation function. Compared with other classic models, the proposed model can not only improve the accuracy of early warning, but also better avoid the problem of gradient disappearance and overfitting.
引文
[1] Zhang Yu-liang.Network opinion risk evaluation index system based on the cycle of emergency[J].Information Science,2012,30(7):1034-1037.(in Chinese)
[2] Tian Yi-lin,Zhou Yan.Network public opinion information monitoring index system model research[M].Berlin:Springer Berlin Heidelberg,2013.
[3] Wen Jin,Hu Liang,Yu Hong-mei.Research on construction of evaluation index system of internet public sentiment for Chinese university [C]//Proc of the 2nd International Conference on Social Science and Higher Education,2016:316-319.
[4] Zhao L Y,Zhang X,He M,et al.Research on public opinion index system of Chinese microblog[C]//Proc of the 5th IEEE International Conference on Software Engineering and Service Science,2014:385-388.
[5] Xu Xue-feng,Du Wan-ying,Hao Xiao-xia,et al.Warning mechanism model of online public opinion based on system dynamics[J].Journal of Qingdao University (Natural Science Edition),2011,24(2),83-86.(in Chinese)
[6] Eirinaki M,Vazirgiannis M.Web mining for personalization[J].ACM Transactions on Internet Technology,2003,3(1):1-27.
[7] Martens D,Bruynseels L,Baesens B,et al.Predicting going concern opinion with data mining[J].Decision Support System,2008,45(4):756-777.
[8] Zhang Yi-wen,Qi Jia-yin,Fang Bin-xing,et al.Online public opinion risk warning based on Bayesian network modeling[J].Library and Information Service,2012,56(2):76-81.(in Chinese)
[9] Hao X H,Gu K C,Meng B Y.Study on the pre-warning system of public opinion based on PCA and PSO-SVM[C]//Proc of the 2016 3rd International Conference on Materials Engineering Manufacturing Technology and Control,2016:661-664.
[10] Zhang Yi-wen,Qi Jia-yin,Fang Bin-xing,et al.Research on the index system of public opinion on internet for abnormal emergency—Concept definition and basic dimensionality[J].Journal of Beijing University of Posts and Telecommunications(Social Sciences Edition),2010,12(4):6-14.(in Chinese)
[11] Wang Qing,Cheng Ying,Chao Nai-peng.On the construction of internet public opinion index system for monitoring and early warning[J].Library and Information Service,2011,55(8):54-57.(in Chinese)
[12] Lan Yue-xin,Deng Xin-yuan,Ma Min.On the construction of internet public opinion evaluating the safety index system of group event[J].Information Research,2011(10):37-39.(in Chinese)
[13] Chen Xin-jie,Hu Yu,Lan Yue-xin.Index systems of network public sentiment for monitoring[J].Modern Information,2012,32(5):3-7.(in Chinese)
[14] Feng Shu-hui.Research on the detection and prediction of food security internet public opinion[D].Xiamen:Jimei University,2015.(in Chinese)
[15] Cao Cheng-yuan,Lü Qiang.Using bidirectional LSTM deep neural network for protein residue contact prediction[J].Journal of Chinese Computer Systems,2017,38(3):531-535.(in Chinese)
[16] Liang Jun,Chai Yu-mei,Yuan Hui-bin,et al.Polarity shifting and LSTM based recursive networks for sentiment analysis[J].Journal of Chinese Information Processing,2015,29(5):152-159.(in Chinese)
[17] Zhao Ming,Du Hui-fang,Dong Cui-cui,et al.Diet health text classification based on word2vec and LSTM[J].Transactions of the Chinese Society of Agricultural,2017,48(10):202-208.(in Chinese)
[18] Kingma D P,Ba J.Adam:A method for stochastic optimization[C]//Proc of the 3rd International Conference for Learning Representations,2015:1-15.附中文参考文献:
[1] 张玉亮.基于发生周期的突发事件网络舆情风险评价指标体系[J].情报科学,2012,30(7):1034-1037.
[5] 徐学峰,杜晚樱,郝晓霞,等.网络舆情预警机制的系统动力学仿真[J].青岛大学学报(自然科学版),2011,24(2):83-86.
[8] 张一文,齐佳音,方滨兴,等.基于贝叶斯网络建模的非常规危机事件网络舆情预警研究[J].图书情报工作,2012,56(2):76-81.
[10] 张一文,齐佳音,方滨兴,等.非常规突发事件网络舆情指标体系建立初探——概念界定与基本维度[J].北京邮电大学学报(社会科学版),2010,12(4):6-14.
[11] 王青,成颖,巢乃鹏.网络舆情监测及预警指标体系构建研究[J].图书情报工作,2011,55(8):54-57.
[12] 兰月新,邓新元,马民.群体性事件网络舆情安全评估指标体系构建[J].情报探索,2011(10):37-39.
[13] 陈新杰,呼雨,兰月新.网络舆情监测指标体系构建研究[J].现代情报,2012,32(5):3-7.
[14] 冯姝慧.食品安全网络舆情监测与预警研究[D].厦门:集美大学,2015.
[15] 曹成远,吕强.使用双向LSTM的深度神经网络预测蛋白质残基相互作用[J].小型微型计算机系统,2017,38(3):531-535.
[16] 梁军,柴玉梅,原慧斌,等.基于极性转移和LSTM递归网络的情感分析[J].中文信息学报,2015,29(5):152-159.
[17] 赵明,杜会芳,董翠翠,等.基于word2vec和LSTM的饮食健康文本分类研究[J].农业机械学报,2017,48(10):202-208.
[2] Tian Yi-lin,Zhou Yan.Network public opinion information monitoring index system model research[M].Berlin:Springer Berlin Heidelberg,2013.
[3] Wen Jin,Hu Liang,Yu Hong-mei.Research on construction of evaluation index system of internet public sentiment for Chinese university [C]//Proc of the 2nd International Conference on Social Science and Higher Education,2016:316-319.
[4] Zhao L Y,Zhang X,He M,et al.Research on public opinion index system of Chinese microblog[C]//Proc of the 5th IEEE International Conference on Software Engineering and Service Science,2014:385-388.
[5] Xu Xue-feng,Du Wan-ying,Hao Xiao-xia,et al.Warning mechanism model of online public opinion based on system dynamics[J].Journal of Qingdao University (Natural Science Edition),2011,24(2),83-86.(in Chinese)
[6] Eirinaki M,Vazirgiannis M.Web mining for personalization[J].ACM Transactions on Internet Technology,2003,3(1):1-27.
[7] Martens D,Bruynseels L,Baesens B,et al.Predicting going concern opinion with data mining[J].Decision Support System,2008,45(4):756-777.
[8] Zhang Yi-wen,Qi Jia-yin,Fang Bin-xing,et al.Online public opinion risk warning based on Bayesian network modeling[J].Library and Information Service,2012,56(2):76-81.(in Chinese)
[9] Hao X H,Gu K C,Meng B Y.Study on the pre-warning system of public opinion based on PCA and PSO-SVM[C]//Proc of the 2016 3rd International Conference on Materials Engineering Manufacturing Technology and Control,2016:661-664.
[10] Zhang Yi-wen,Qi Jia-yin,Fang Bin-xing,et al.Research on the index system of public opinion on internet for abnormal emergency—Concept definition and basic dimensionality[J].Journal of Beijing University of Posts and Telecommunications(Social Sciences Edition),2010,12(4):6-14.(in Chinese)
[11] Wang Qing,Cheng Ying,Chao Nai-peng.On the construction of internet public opinion index system for monitoring and early warning[J].Library and Information Service,2011,55(8):54-57.(in Chinese)
[12] Lan Yue-xin,Deng Xin-yuan,Ma Min.On the construction of internet public opinion evaluating the safety index system of group event[J].Information Research,2011(10):37-39.(in Chinese)
[13] Chen Xin-jie,Hu Yu,Lan Yue-xin.Index systems of network public sentiment for monitoring[J].Modern Information,2012,32(5):3-7.(in Chinese)
[14] Feng Shu-hui.Research on the detection and prediction of food security internet public opinion[D].Xiamen:Jimei University,2015.(in Chinese)
[15] Cao Cheng-yuan,Lü Qiang.Using bidirectional LSTM deep neural network for protein residue contact prediction[J].Journal of Chinese Computer Systems,2017,38(3):531-535.(in Chinese)
[16] Liang Jun,Chai Yu-mei,Yuan Hui-bin,et al.Polarity shifting and LSTM based recursive networks for sentiment analysis[J].Journal of Chinese Information Processing,2015,29(5):152-159.(in Chinese)
[17] Zhao Ming,Du Hui-fang,Dong Cui-cui,et al.Diet health text classification based on word2vec and LSTM[J].Transactions of the Chinese Society of Agricultural,2017,48(10):202-208.(in Chinese)
[18] Kingma D P,Ba J.Adam:A method for stochastic optimization[C]//Proc of the 3rd International Conference for Learning Representations,2015:1-15.附中文参考文献:
[1] 张玉亮.基于发生周期的突发事件网络舆情风险评价指标体系[J].情报科学,2012,30(7):1034-1037.
[5] 徐学峰,杜晚樱,郝晓霞,等.网络舆情预警机制的系统动力学仿真[J].青岛大学学报(自然科学版),2011,24(2):83-86.
[8] 张一文,齐佳音,方滨兴,等.基于贝叶斯网络建模的非常规危机事件网络舆情预警研究[J].图书情报工作,2012,56(2):76-81.
[10] 张一文,齐佳音,方滨兴,等.非常规突发事件网络舆情指标体系建立初探——概念界定与基本维度[J].北京邮电大学学报(社会科学版),2010,12(4):6-14.
[11] 王青,成颖,巢乃鹏.网络舆情监测及预警指标体系构建研究[J].图书情报工作,2011,55(8):54-57.
[12] 兰月新,邓新元,马民.群体性事件网络舆情安全评估指标体系构建[J].情报探索,2011(10):37-39.
[13] 陈新杰,呼雨,兰月新.网络舆情监测指标体系构建研究[J].现代情报,2012,32(5):3-7.
[14] 冯姝慧.食品安全网络舆情监测与预警研究[D].厦门:集美大学,2015.
[15] 曹成远,吕强.使用双向LSTM的深度神经网络预测蛋白质残基相互作用[J].小型微型计算机系统,2017,38(3):531-535.
[16] 梁军,柴玉梅,原慧斌,等.基于极性转移和LSTM递归网络的情感分析[J].中文信息学报,2015,29(5):152-159.
[17] 赵明,杜会芳,董翠翠,等.基于word2vec和LSTM的饮食健康文本分类研究[J].农业机械学报,2017,48(10):202-208.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |