Autoencoder with improved SPNs and its application in sentiment analysis for short texts

被引：0
|
作者：
Wang S. [1 ]
Zhang H. [2 ]
Pan Y. [1 ]
机构：
[1] College of Computer Science and Technology, Jilin University, Changchun
[2] College of Software, Jilin University, Changchun
来源：
Harbin Gongcheng Daxue Xuebao/Journal of Harbin Engineering University | 2020年 / 41卷 / 03期
关键词：
Deep autoencoder; Deep learning; Feature extraction; Online learning; Sentiment analysis; Structure learning; Sum-product networks model; Tractable model;
D O I：
10.11990/jheu.201808087
中图分类号：
学科分类号：
摘要：
As one of the main components of deep learning structure, the deep autoencoder plays an important role in both unsupervised learning and non-linear feature extraction, which have made significant progress in natural language processing and other fields. To improve the performance of deep autoencoder in sentiment analysis application of short texts, a deep autoencoder based on improved sum-product networks (SPNs) is proposed. First, the node layers of Spns is reconstructed, the input layer's output is added to each hidden layer in Spns, a layered Spns is proposed and the deep autoencoder based on Lspns is constructed. At the same time, a max-Product Networks model that changes Spns' Sum node to max node is proposed as a decoder. Finally, the Lspns-based deep autoencoder is applied to short text sentiment analysis. Experiment result shows that Lspns-based deep autoencoder can obtain higher classification accuracy and run faster in the short text sentiment analysis field than the existing deep autoencoder. © 2020, Editorial Department of Journal of HEU. All right reserved.
引用
收藏
页码：411 / 419
页数：8
相关论文
共 21 条
  • [1] BENGIO Y, YAO Li, ALAIN G, Et al., Generalized denoising auto-encoders as generative models, Proceedings of the 26th International Conference on Neural Information Processing Systems, pp. 899-907, (2013)
  • [2] HINTON G E, OSINDERO S, TEH Y W., A fast learning algorithm for deep belief nets, Neural computation, 18, 7, pp. 1527-1554, (2006)
  • [3] POON H, DOMINGOS P., Sum-product networks: a new deep architecture, Proceedings of 2011 IEEE International Conference on Computer Vision Workshops, pp. 337-346, (2011)
  • [4] ZHAO Han, POUPART P, GORDON G., A unified approach for learning the parameters of sum-product networks, Proceedings of the 30th International Conference on Neural Information Processing Systems, pp. 433-441, (2016)
  • [5] GENS R, DOMINGOS P., Discriminative learning of sum-product networks, Proceedings of the 25th International Conference on Neural Information Processing Systems, pp. 3239-3247, (2012)
  • [6] CHENG Weichen, KOK S, PHAM H V, Et al., Language modeling with sum-product networks, Proceedings of the 15th Annual Conference of the International Speech Communication Association, pp. 2098-2102, (2014)
  • [7] AMER M R, TODOROVIC S., Sum product networks for activity recognition, IEEE transactions on pattern analysis and machine intelligence, 38, 4, pp. 800-813, (2016)
  • [8] GENS R, DOMINGOS P., Learning the structure of sum-product networks, Proceedings of the 30th International Conference on Machine Learning, pp. 873-880, (2013)
  • [9] MOLINA A, NATARAJAN S, KERSTING K., Poisson sum-product networks: a deep architecture for tractable multivariate Poisson distributions, Proceedings of the 31st AAAI Conference on Artificial Intelligence, pp. 2357-2363, (2017)
  • [10] TRAPP M, MADL T, PEHARZ R, Et al., Safe semi-supervised learning of sum-product networks, Proceedings of the 33rd Conference on Uncertainty in Artificial Intelligence, pp. 143-149, (2017)
123