An Ensemble Neural Network Model for Benefiting Pregnancy Health Stats from Mining Social Media

Neha Warikoo, Yung Chun Chang, Hong Jie Dai, Wen Lian Hsu

研究成果: 書貢獻/報告類型會議貢獻

摘要

Extensive use of social media for communication has made it a desired resource in human behavior intensive tasks like product popularity, public polls and more recently for public health surveillance tasks such as lifestyle associated diseases and mental health. In this paper, we exploited Twitter data for detecting pregnancy cases and used tweets about pregnancy to study trigger terms associated with maternal physical and mental health. Such systems can enable clinicians to offer a more comprehensive health care in real time. Using a Twitter-based corpus, we have developed an ensemble Long-short Term Memory (LSTM) – Recurrent Neural Networks (RNN) and Convolution Neural Networks (CNN) network representation model to learn legitimate pregnancy cases discussed online. These ensemble representations were learned by a SVM classifier, which can achieve F1-score of 95% in predicting pregnancy accounts discussed in tweets. We also further investigate the words most commonly associated with physical disease symptoms ‘Distress’ and negative emotions ‘Annoyed’ sentiment. Results from our sentiment analysis study are quite encouraging, identifying more accurate triggers for pregnancy sentiment classes.

原文英語
主出版物標題Information Retrieval Technology - 14th Asia Information Retrieval Societies Conference, AIRS 2018, Proceedings
編輯Lun-Wei Ku, Jui-Feng Yeh, Liang-Chih Yu, Yuen-Hsien Tseng, Zhi-Hong Chen, Tetsuya Sakai, Jing Jiang, Lung-Hao Lee, Dae Hoon Park
發行者Springer Verlag
頁面3-15
頁數13
ISBN(列印)9783030035198
DOIs
出版狀態已發佈 - 一月 1 2018
事件14th Asia Information Retrieval Societies conference, AIRS 2018 - Taipei, 臺灣
持續時間: 十一月 28 2018十一月 30 2018

出版系列

名字Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
11292 LNCS
ISSN(列印)0302-9743
ISSN(電子)1611-3349

會議

會議14th Asia Information Retrieval Societies conference, AIRS 2018
國家臺灣
城市Taipei
期間11/28/1811/30/18

指紋

Social Media
Pregnancy
Neural Network Model
Mining
Health
Ensemble
Neural networks
Recurrent neural networks
Public health
Convolution
Health care
Classifiers
Trigger
Communication
Sentiment Analysis
Memory Term
Human Behavior
Public Health
Recurrent Neural Networks
Surveillance

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

引用此文

Warikoo, N., Chang, Y. C., Dai, H. J., & Hsu, W. L. (2018). An Ensemble Neural Network Model for Benefiting Pregnancy Health Stats from Mining Social Media. 於 L-W. Ku, J-F. Yeh, L-C. Yu, Y-H. Tseng, Z-H. Chen, T. Sakai, J. Jiang, L-H. Lee, ... D. H. Park (編輯), Information Retrieval Technology - 14th Asia Information Retrieval Societies Conference, AIRS 2018, Proceedings (頁 3-15). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); 卷 11292 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-03520-4_1

An Ensemble Neural Network Model for Benefiting Pregnancy Health Stats from Mining Social Media. / Warikoo, Neha; Chang, Yung Chun; Dai, Hong Jie; Hsu, Wen Lian.

Information Retrieval Technology - 14th Asia Information Retrieval Societies Conference, AIRS 2018, Proceedings. 編輯 / Lun-Wei Ku; Jui-Feng Yeh; Liang-Chih Yu; Yuen-Hsien Tseng; Zhi-Hong Chen; Tetsuya Sakai; Jing Jiang; Lung-Hao Lee; Dae Hoon Park. Springer Verlag, 2018. p. 3-15 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); 卷 11292 LNCS).

研究成果: 書貢獻/報告類型會議貢獻

Warikoo, N, Chang, YC, Dai, HJ & Hsu, WL 2018, An Ensemble Neural Network Model for Benefiting Pregnancy Health Stats from Mining Social Media. 於 L-W Ku, J-F Yeh, L-C Yu, Y-H Tseng, Z-H Chen, T Sakai, J Jiang, L-H Lee & DH Park (編輯), Information Retrieval Technology - 14th Asia Information Retrieval Societies Conference, AIRS 2018, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 卷 11292 LNCS, Springer Verlag, 頁 3-15, 14th Asia Information Retrieval Societies conference, AIRS 2018, Taipei, 臺灣, 11/28/18. https://doi.org/10.1007/978-3-030-03520-4_1
Warikoo N, Chang YC, Dai HJ, Hsu WL. An Ensemble Neural Network Model for Benefiting Pregnancy Health Stats from Mining Social Media. 於 Ku L-W, Yeh J-F, Yu L-C, Tseng Y-H, Chen Z-H, Sakai T, Jiang J, Lee L-H, Park DH, 編輯, Information Retrieval Technology - 14th Asia Information Retrieval Societies Conference, AIRS 2018, Proceedings. Springer Verlag. 2018. p. 3-15. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-030-03520-4_1
Warikoo, Neha ; Chang, Yung Chun ; Dai, Hong Jie ; Hsu, Wen Lian. / An Ensemble Neural Network Model for Benefiting Pregnancy Health Stats from Mining Social Media. Information Retrieval Technology - 14th Asia Information Retrieval Societies Conference, AIRS 2018, Proceedings. 編輯 / Lun-Wei Ku ; Jui-Feng Yeh ; Liang-Chih Yu ; Yuen-Hsien Tseng ; Zhi-Hong Chen ; Tetsuya Sakai ; Jing Jiang ; Lung-Hao Lee ; Dae Hoon Park. Springer Verlag, 2018. 頁 3-15 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{b479627b08dc48cabb1ed9e6eed51831,
title = "An Ensemble Neural Network Model for Benefiting Pregnancy Health Stats from Mining Social Media",
abstract = "Extensive use of social media for communication has made it a desired resource in human behavior intensive tasks like product popularity, public polls and more recently for public health surveillance tasks such as lifestyle associated diseases and mental health. In this paper, we exploited Twitter data for detecting pregnancy cases and used tweets about pregnancy to study trigger terms associated with maternal physical and mental health. Such systems can enable clinicians to offer a more comprehensive health care in real time. Using a Twitter-based corpus, we have developed an ensemble Long-short Term Memory (LSTM) – Recurrent Neural Networks (RNN) and Convolution Neural Networks (CNN) network representation model to learn legitimate pregnancy cases discussed online. These ensemble representations were learned by a SVM classifier, which can achieve F1-score of 95{\%} in predicting pregnancy accounts discussed in tweets. We also further investigate the words most commonly associated with physical disease symptoms ‘Distress’ and negative emotions ‘Annoyed’ sentiment. Results from our sentiment analysis study are quite encouraging, identifying more accurate triggers for pregnancy sentiment classes.",
keywords = "Ensemble deep learning, Health surveillance, Pregnancy health stats, Sentiment analysis, Text mining of Twitter data",
author = "Neha Warikoo and Chang, {Yung Chun} and Dai, {Hong Jie} and Hsu, {Wen Lian}",
year = "2018",
month = "1",
day = "1",
doi = "10.1007/978-3-030-03520-4_1",
language = "English",
isbn = "9783030035198",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "3--15",
editor = "Lun-Wei Ku and Jui-Feng Yeh and Liang-Chih Yu and Yuen-Hsien Tseng and Zhi-Hong Chen and Tetsuya Sakai and Jing Jiang and Lung-Hao Lee and Park, {Dae Hoon}",
booktitle = "Information Retrieval Technology - 14th Asia Information Retrieval Societies Conference, AIRS 2018, Proceedings",
address = "Germany",

}

TY - GEN

T1 - An Ensemble Neural Network Model for Benefiting Pregnancy Health Stats from Mining Social Media

AU - Warikoo, Neha

AU - Chang, Yung Chun

AU - Dai, Hong Jie

AU - Hsu, Wen Lian

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Extensive use of social media for communication has made it a desired resource in human behavior intensive tasks like product popularity, public polls and more recently for public health surveillance tasks such as lifestyle associated diseases and mental health. In this paper, we exploited Twitter data for detecting pregnancy cases and used tweets about pregnancy to study trigger terms associated with maternal physical and mental health. Such systems can enable clinicians to offer a more comprehensive health care in real time. Using a Twitter-based corpus, we have developed an ensemble Long-short Term Memory (LSTM) – Recurrent Neural Networks (RNN) and Convolution Neural Networks (CNN) network representation model to learn legitimate pregnancy cases discussed online. These ensemble representations were learned by a SVM classifier, which can achieve F1-score of 95% in predicting pregnancy accounts discussed in tweets. We also further investigate the words most commonly associated with physical disease symptoms ‘Distress’ and negative emotions ‘Annoyed’ sentiment. Results from our sentiment analysis study are quite encouraging, identifying more accurate triggers for pregnancy sentiment classes.

AB - Extensive use of social media for communication has made it a desired resource in human behavior intensive tasks like product popularity, public polls and more recently for public health surveillance tasks such as lifestyle associated diseases and mental health. In this paper, we exploited Twitter data for detecting pregnancy cases and used tweets about pregnancy to study trigger terms associated with maternal physical and mental health. Such systems can enable clinicians to offer a more comprehensive health care in real time. Using a Twitter-based corpus, we have developed an ensemble Long-short Term Memory (LSTM) – Recurrent Neural Networks (RNN) and Convolution Neural Networks (CNN) network representation model to learn legitimate pregnancy cases discussed online. These ensemble representations were learned by a SVM classifier, which can achieve F1-score of 95% in predicting pregnancy accounts discussed in tweets. We also further investigate the words most commonly associated with physical disease symptoms ‘Distress’ and negative emotions ‘Annoyed’ sentiment. Results from our sentiment analysis study are quite encouraging, identifying more accurate triggers for pregnancy sentiment classes.

KW - Ensemble deep learning

KW - Health surveillance

KW - Pregnancy health stats

KW - Sentiment analysis

KW - Text mining of Twitter data

UR - http://www.scopus.com/inward/record.url?scp=85057269768&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85057269768&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-03520-4_1

DO - 10.1007/978-3-030-03520-4_1

M3 - Conference contribution

AN - SCOPUS:85057269768

SN - 9783030035198

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 3

EP - 15

BT - Information Retrieval Technology - 14th Asia Information Retrieval Societies Conference, AIRS 2018, Proceedings

A2 - Ku, Lun-Wei

A2 - Yeh, Jui-Feng

A2 - Yu, Liang-Chih

A2 - Tseng, Yuen-Hsien

A2 - Chen, Zhi-Hong

A2 - Sakai, Tetsuya

A2 - Jiang, Jing

A2 - Lee, Lung-Hao

A2 - Park, Dae Hoon

PB - Springer Verlag

ER -