Advertisement

Text-Visualizing Neural Network Model: Understanding Online Financial Textual Data

  • Tomoki ItoEmail author
  • Hiroki Sakaji
  • Kota Tsubouchi
  • Kiyoshi Izumi
  • Tatsuo Yamashita
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 10939)

Abstract

This study aims to visualize financial documents to swiftly obtain market sentiment information from these documents and determine the reason for which sentiment decisions are made. This type of visualization is considered helpful for nonexperts to easily understand technical documents such as financial reports. To achieve this, we propose a novel interpretable neural network (NN) architecture called gradient interpretable NN (GINN). GINN can visualize both the market sentiment score from a whole financial document and the sentiment gradient scores in concept units. We experimentally demonstrate the validity of text visualization produced by GINN using a real textual dataset.

Keywords

Interpretable neural network Text mining Support system 

Notes

Acknowledgment

This work was supported in part by JSPS Fellows Grant Number 17J04768.

Supplementary material

469284_1_En_20_MOESM1_ESM.pdf (133 kb)
Supplementary material 1 (pdf 133 KB)

References

  1. 1.
    Ravi, K., Ravi, V.: A survey on opinion mining and sentiment analysis: tasks, approaches and applications. Knowl.-Based Syst. 89(C), 14–46 (2015)CrossRefGoogle Scholar
  2. 2.
    Hechtlinger, Y.: Interpretation of prediction models using the input gradient. In: NIPS 2016 Workshop on Interpretable Machine Learning in Complex Systems (2016)Google Scholar
  3. 3.
    Bach, S., Binder, A., Montavon, G., Klauschen, F., Muller, K.R., Samek, W.: On pixel-wise explanations for nonlinear classifier decisions by layer-wise relevance propagation. PLoS ONE 10(7), 1–46 (2015)Google Scholar
  4. 4.
    Mikolov, T., Chen, K., Sutskever, I., Corrado, G., Dean, J.: Distributed representations of words and phrases and their compositionality. In: NIPS 2013, pp. 3111–3119 (2013)Google Scholar
  5. 5.
    Hornik, K., Feinerer, I., Kober, M., Buchta, C.: Spherical k-means clustering. J. Stat. Softw. 50(10), 1–22 (2012)CrossRefGoogle Scholar
  6. 6.
    Zhao, P., Zhang, T.: Accelerating minibatch stochastic gradient descent using stratified sampling. arXiv:1405.3080v1 (2014)
  7. 7.
    Kingma, D.P., Ba., J.L.: Adam: a method for stochastic optimization. In: ICLR 2015 (2015)Google Scholar
  8. 8.
    Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I., Salakhutdinov, R.: Dropout: a simple way to prevent neural networks from overfitting. J. Mach. Learn. Res. 15(1), 1929–1958 (2014)MathSciNetzbMATHGoogle Scholar
  9. 9.
    Kjellin, P.E., Liu, Y.: A survey on interactivity in topic models. IJACSA 7(14), 456–461 (2016)Google Scholar
  10. 10.
    Tandem anchoring: a multiword anchor approach for interactive topic modeling. In: ACL 2017, pp. 896–905 (2017)Google Scholar
  11. 11.
    Shrikumar, A., Greenside, P., Kundaje, A.: Learning important features through propagating activation differences. In: ICML 2017 (2017)Google Scholar
  12. 12.
    Xu, Q., Zhao, Q., Pei, W., Yang, L., He, Z.: Design interpretable neural network trees through self-organized learning of features. In: IJCNN 2004 (2004)Google Scholar
  13. 13.
    Zhang, Q., Wu, Y.N., Zhu, S.: Interpretable convolutional neural networks. arXiv:1710.00935 (2017)
  14. 14.
    Mnih, V., Heess, N., Graves, A., Kavukcuoglu, K.: Recurrent models of visual attention. In: NIPS 2014, pp. 2204–2212 (2014)Google Scholar
  15. 15.
    Xu, K., Ba, J., Kiros, R., Cho, K., Courville, A., Salakhutdinov, R., Zemel, R., Bengio, Y.: Show, attend and tell: neural image caption generation with visual attention. In: ICML 2015, pp. 77–81 (2015)Google Scholar
  16. 16.
    Dong, Y., Su, H., Zhu., J, Zhang, B.: Improving interpretability of deep neural networks with semantic information. In: CVPR 2017 (2017)Google Scholar

Copyright information

© Springer International Publishing AG, part of Springer Nature 2018

Authors and Affiliations

  • Tomoki Ito
    • 1
    Email author
  • Hiroki Sakaji
    • 1
  • Kota Tsubouchi
    • 2
  • Kiyoshi Izumi
    • 1
  • Tatsuo Yamashita
    • 2
  1. 1.The University of TokyoTokyoJapan
  2. 2.Yahoo Japan CorporationTokyoJapan

Personalised recommendations