5 Conclusion and Future Works
In this paper, the RCNN model is used to identify emergencies. Due to the shortcomings of traditional RCNN models in event recognition, this paper tries to improve
and optimize the traditional models by combining different RNN models and activation
functions. At the same time, we adjust the data scale and word vector dimension to get
the best model.
The next step will accept a strategy to increase the speed of training, reduce the
time of training, and combine with web crawlers to create an online emergency
identification system to provide an external user with a continuous and reliable service.
Acknowledgements. This work is supported by Xinjiang Joint Fund of National Science Fund
of China (U1703261). We are grateful to Shanghai University, Sina News for their provided open
resources.
References
1. Phillips BD, Neal DM, Webb G (2016) Introduction to emergency management. CRC Press
2. Qin X, Zu-Jun M, Hua-Jun L (2008) Location-routing problem in emergency logistics for
public emergencies. J Huazhong Univ Sci Technol
3. Yankova M, Boytcheva S (2003) Focusing on scenario recognition in information
extraction. In: Tenth conference on European chapter of the association for computational
linguistics. Association for Computational Linguistics
4. Rainer (2007) Modeling treatment processes using information extraction. Biochem Soc
Trans 38(6):1581–1586
5. Yangarber R (2001) Scenario customization for information extraction. Dissertations &
Theses
6. Abadi M, Barham P, Chen J et al (2016) TensorFlow: a system for large-scale machine
learning
7. Ahn DD (2006) The stages of event extraction. Workshop on annotating & reasoning about
time & events. Association for Computational Linguistics
8. Vlachos A, Craven M (2012) Biomedical event extraction from abstracts and full papers
using search-based structured prediction. BMC Bioinform 13(11 suppl):S5
9. Lecun Y, Bengio Y, Hinton G (2015) Deep learning. Nature 521(7553):436–444
10. Kim Y (2014) Convolutional neural networks for sentence classification. Eprint Arxiv
11. Su B, Zhang X, Lu S et al (2015) Segmented handwritten text recognition with recurrent
neural network classifiers. In: 2015 13th international conference on document analysis and
recognition (ICDAR). IEEE Computer Society
12. Lai S, Xu L, Liu K et al (2015) Recurrent convolutional neural networks for text
classification. In: Twenty-ninth AAAI conference on artificial intelligence
104
H. He et al.
In this paper, the RCNN model is used to identify emergencies. Due to the shortcomings of traditional RCNN models in event recognition, this paper tries to improve
and optimize the traditional models by combining different RNN models and activation
functions. At the same time, we adjust the data scale and word vector dimension to get
the best model.
The next step will accept a strategy to increase the speed of training, reduce the
time of training, and combine with web crawlers to create an online emergency
identification system to provide an external user with a continuous and reliable service.
Acknowledgements. This work is supported by Xinjiang Joint Fund of National Science Fund
of China (U1703261). We are grateful to Shanghai University, Sina News for their provided open
resources.
References
1. Phillips BD, Neal DM, Webb G (2016) Introduction to emergency management. CRC Press
2. Qin X, Zu-Jun M, Hua-Jun L (2008) Location-routing problem in emergency logistics for
public emergencies. J Huazhong Univ Sci Technol
3. Yankova M, Boytcheva S (2003) Focusing on scenario recognition in information
extraction. In: Tenth conference on European chapter of the association for computational
linguistics. Association for Computational Linguistics
4. Rainer (2007) Modeling treatment processes using information extraction. Biochem Soc
Trans 38(6):1581–1586
5. Yangarber R (2001) Scenario customization for information extraction. Dissertations &
Theses
6. Abadi M, Barham P, Chen J et al (2016) TensorFlow: a system for large-scale machine
learning
7. Ahn DD (2006) The stages of event extraction. Workshop on annotating & reasoning about
time & events. Association for Computational Linguistics
8. Vlachos A, Craven M (2012) Biomedical event extraction from abstracts and full papers
using search-based structured prediction. BMC Bioinform 13(11 suppl):S5
9. Lecun Y, Bengio Y, Hinton G (2015) Deep learning. Nature 521(7553):436–444
10. Kim Y (2014) Convolutional neural networks for sentence classification. Eprint Arxiv
11. Su B, Zhang X, Lu S et al (2015) Segmented handwritten text recognition with recurrent
neural network classifiers. In: 2015 13th international conference on document analysis and
recognition (ICDAR). IEEE Computer Society
12. Lai S, Xu L, Liu K et al (2015) Recurrent convolutional neural networks for text
classification. In: Twenty-ninth AAAI conference on artificial intelligence
104
H. He et al.