Identifying Emergency Stages in Facebook Posts of Police Departments with Convolutional and Recurrent Neural Networks and Support Vector Machines

Nicolai Pogrebnyakov, Edgar Maldonado

    Publikation: Kapitel i bog/rapport/konferenceprocesKonferencebidrag i proceedingsForskningpeer review

    Resumé

    Classification of social media posts in emergency response is an important practical problem: accurate classification can help automate processing of such messages and help other responders and the public react to emergencies in a timely fashion. This research focused on classifying Facebook messages of US police departments. Randomly selected 5,000 messages were used to train classifiers that distinguished between four categories of messages: emergency preparedness, response and recovery, as well as general engagement messages. Features were represented with bag-of-words and word2vec, and models were constructed using support vector machines (SVMs) and convolutional (CNNs) and recurrent neural networks (RNNs). The best performing classifier was an RNN with a custom-trained word2vec model to represent features, which achieved the F1 measure of 0.839.
    Classification of social media posts in emergency response is an important practical problem: accurate classification can help automate processing of such messages and help other responders and the public react to emergencies in a timely fashion. This research focused on classifying Facebook messages of US police departments. Randomly selected 5,000 messages were used to train classifiers that distinguished between four categories of messages: emergency preparedness, response and recovery, as well as general engagement messages. Features were represented with bag-of-words and word2vec, and models were constructed using support vector machines (SVMs) and convolutional (CNNs) and recurrent neural networks (RNNs). The best performing classifier was an RNN with a custom-trained word2vec model to represent features, which achieved the F1 measure of 0.839.
    SprogEngelsk
    TitelProceedings. 2017 IEEE International Conference on Big Data : IEEE Big Data 2017
    RedaktørerJian-Yun Nie, Zoran Obradovic, Toyotaro Suzumura, Rumi Ghosh, Raghunath Nambiar, Chonggang Wang, Hui Zang, Ricardo Baeza-Yates, Xiaohua Hu, Jeremy Kepner, Alfredo Cuzzocrea, Jian Tang, Masashi Toyoda
    Udgivelses stedLos Alamitos, CA
    ForlagIEEE
    Dato2017
    Sider4343-4352
    ISBN (Trykt)9781538627167
    ISBN (Elektronisk)9781538627150, 9781538627143
    DOI
    StatusUdgivet - 2017
    Begivenhed5th IEEE International Conference on Big Data. 2017 - Boston, USA
    Varighed: 11 dec. 201714 dec. 2017
    Konferencens nummer: 5
    http://cci.drexel.edu/bigdata/bigdata2017/

    Konference

    Konference5th IEEE International Conference on Big Data. 2017
    Nummer5
    LandUSA
    ByBoston
    Periode11/12/201714/12/2017
    Internetadresse

    Bibliografisk note

    CBS Bibliotek har ikke adgang til materialet

    Emneord

    • Social media
    • Classification
    • Police
    • Support vector machines
    • Neural networks

    Citer dette

    Pogrebnyakov, N., & Maldonado, E. (2017). Identifying Emergency Stages in Facebook Posts of Police Departments with Convolutional and Recurrent Neural Networks and Support Vector Machines. I J-Y. Nie, Z. Obradovic, T. Suzumura, R. Ghosh, R. Nambiar, C. Wang, H. Zang, R. Baeza-Yates, X. Hu, J. Kepner, A. Cuzzocrea, J. Tang, ... M. Toyoda (red.), Proceedings. 2017 IEEE International Conference on Big Data: IEEE Big Data 2017 (s. 4343-4352). Los Alamitos, CA: IEEE. DOI: 10.1109/BigData.2017.8258464
    Pogrebnyakov, Nicolai ; Maldonado, Edgar. / Identifying Emergency Stages in Facebook Posts of Police Departments with Convolutional and Recurrent Neural Networks and Support Vector Machines. Proceedings. 2017 IEEE International Conference on Big Data: IEEE Big Data 2017. red. / Jian-Yun Nie ; Zoran Obradovic ; Toyotaro Suzumura ; Rumi Ghosh ; Raghunath Nambiar ; Chonggang Wang ; Hui Zang ; Ricardo Baeza-Yates ; Xiaohua Hu ; Jeremy Kepner ; Alfredo Cuzzocrea ; Jian Tang ; Masashi Toyoda. Los Alamitos, CA : IEEE, 2017. s. 4343-4352
    @inproceedings{b02d47fdb20f45c1a8cfa394be75bf78,
    title = "Identifying Emergency Stages in Facebook Posts of Police Departments with Convolutional and Recurrent Neural Networks and Support Vector Machines",
    abstract = "Classification of social media posts in emergency response is an important practical problem: accurate classification can help automate processing of such messages and help other responders and the public react to emergencies in a timely fashion. This research focused on classifying Facebook messages of US police departments. Randomly selected 5,000 messages were used to train classifiers that distinguished between four categories of messages: emergency preparedness, response and recovery, as well as general engagement messages. Features were represented with bag-of-words and word2vec, and models were constructed using support vector machines (SVMs) and convolutional (CNNs) and recurrent neural networks (RNNs). The best performing classifier was an RNN with a custom-trained word2vec model to represent features, which achieved the F1 measure of 0.839.",
    keywords = "Social media, Classification, Police, Support vector machines, Neural networks, Social media, Classification, Police, Support vector machines, Neural networks",
    author = "Nicolai Pogrebnyakov and Edgar Maldonado",
    note = "CBS Library does not have access to the material",
    year = "2017",
    doi = "10.1109/BigData.2017.8258464",
    language = "English",
    isbn = "9781538627167",
    pages = "4343--4352",
    editor = "Jian-Yun Nie and Zoran Obradovic and Toyotaro Suzumura and Rumi Ghosh and Raghunath Nambiar and Chonggang Wang and Hui Zang and Ricardo Baeza-Yates and Xiaohua Hu and Jeremy Kepner and Alfredo Cuzzocrea and Jian Tang and Masashi Toyoda",
    booktitle = "Proceedings. 2017 IEEE International Conference on Big Data",
    publisher = "IEEE",
    address = "United States",

    }

    Pogrebnyakov, N & Maldonado, E 2017, Identifying Emergency Stages in Facebook Posts of Police Departments with Convolutional and Recurrent Neural Networks and Support Vector Machines. i J-Y Nie, Z Obradovic, T Suzumura, R Ghosh, R Nambiar, C Wang, H Zang, R Baeza-Yates, X Hu, J Kepner, A Cuzzocrea, J Tang & M Toyoda (red), Proceedings. 2017 IEEE International Conference on Big Data: IEEE Big Data 2017. IEEE, Los Alamitos, CA, s. 4343-4352, 5th IEEE International Conference on Big Data. 2017, Boston, USA, 11/12/2017. DOI: 10.1109/BigData.2017.8258464

    Identifying Emergency Stages in Facebook Posts of Police Departments with Convolutional and Recurrent Neural Networks and Support Vector Machines. / Pogrebnyakov, Nicolai; Maldonado, Edgar.

    Proceedings. 2017 IEEE International Conference on Big Data: IEEE Big Data 2017. red. / Jian-Yun Nie; Zoran Obradovic; Toyotaro Suzumura; Rumi Ghosh; Raghunath Nambiar; Chonggang Wang; Hui Zang; Ricardo Baeza-Yates; Xiaohua Hu; Jeremy Kepner; Alfredo Cuzzocrea; Jian Tang; Masashi Toyoda. Los Alamitos, CA : IEEE, 2017. s. 4343-4352.

    Publikation: Kapitel i bog/rapport/konferenceprocesKonferencebidrag i proceedingsForskningpeer review

    TY - GEN

    T1 - Identifying Emergency Stages in Facebook Posts of Police Departments with Convolutional and Recurrent Neural Networks and Support Vector Machines

    AU - Pogrebnyakov,Nicolai

    AU - Maldonado,Edgar

    N1 - CBS Library does not have access to the material

    PY - 2017

    Y1 - 2017

    N2 - Classification of social media posts in emergency response is an important practical problem: accurate classification can help automate processing of such messages and help other responders and the public react to emergencies in a timely fashion. This research focused on classifying Facebook messages of US police departments. Randomly selected 5,000 messages were used to train classifiers that distinguished between four categories of messages: emergency preparedness, response and recovery, as well as general engagement messages. Features were represented with bag-of-words and word2vec, and models were constructed using support vector machines (SVMs) and convolutional (CNNs) and recurrent neural networks (RNNs). The best performing classifier was an RNN with a custom-trained word2vec model to represent features, which achieved the F1 measure of 0.839.

    AB - Classification of social media posts in emergency response is an important practical problem: accurate classification can help automate processing of such messages and help other responders and the public react to emergencies in a timely fashion. This research focused on classifying Facebook messages of US police departments. Randomly selected 5,000 messages were used to train classifiers that distinguished between four categories of messages: emergency preparedness, response and recovery, as well as general engagement messages. Features were represented with bag-of-words and word2vec, and models were constructed using support vector machines (SVMs) and convolutional (CNNs) and recurrent neural networks (RNNs). The best performing classifier was an RNN with a custom-trained word2vec model to represent features, which achieved the F1 measure of 0.839.

    KW - Social media

    KW - Classification

    KW - Police

    KW - Support vector machines

    KW - Neural networks

    KW - Social media

    KW - Classification

    KW - Police

    KW - Support vector machines

    KW - Neural networks

    U2 - 10.1109/BigData.2017.8258464

    DO - 10.1109/BigData.2017.8258464

    M3 - Article in proceedings

    SN - 9781538627167

    SP - 4343

    EP - 4352

    BT - Proceedings. 2017 IEEE International Conference on Big Data

    PB - IEEE

    CY - Los Alamitos, CA

    ER -

    Pogrebnyakov N, Maldonado E. Identifying Emergency Stages in Facebook Posts of Police Departments with Convolutional and Recurrent Neural Networks and Support Vector Machines. I Nie J-Y, Obradovic Z, Suzumura T, Ghosh R, Nambiar R, Wang C, Zang H, Baeza-Yates R, Hu X, Kepner J, Cuzzocrea A, Tang J, Toyoda M, red., Proceedings. 2017 IEEE International Conference on Big Data: IEEE Big Data 2017. Los Alamitos, CA: IEEE. 2017. s. 4343-4352. Tilgængelig fra, DOI: 10.1109/BigData.2017.8258464