Deep Sentiment Classification and Topic Discovery on Novel Coronavirus or COVID-19 Online Discussions: NLP Using LSTM Recurrent Neural Network Approach

Internet forums and public social media, such as online healthcare forums, provide a convenient channel for users (people/patients) concerned about health issues to discuss and share information with each other. In late December 2019, an outbreak of a novel coronavirus (infection from which results in the disease named COVID-19) was reported, and, due to the rapid spread of the virus in other parts of the world, the World Health Organization declared a state of emergency. In this paper, we used automated extraction of COVID-19–related discussions from social media and a natural language process (NLP) method based on topic modeling to uncover various issues related to COVID-19 from public opinions. Moreover, we also investigate how to use LSTM recurrent neural network for sentiment classification of COVID-19 comments. Our findings shed light on the importance of using public opinions and suitable computational techniques to understand issues surrounding COVID-19 and to guide related decision-making. In addition, experiments demonstrated that the research model achieved an accuracy of 81.15% – a higher accuracy than that of several other well-known machine-learning algorithms for COVID-19–Sentiment Classification.

[1]  J. Lipoff,et al.  Natural Language Processing of Reddit Data to Evaluate Dermatology Patient Experiences and Therapeutics. , 2020, Journal of the American Academy of Dermatology.

[2]  Ge Yu,et al.  Sentiment analysis using deep learning approaches: an overview , 2019, Science China Information Sciences.

[3]  Jonathan Gemmell,et al.  Detecting and Characterizing Trends in Online Mental Health Discussions , 2018, 2018 IEEE International Conference on Data Mining Workshops (ICDMW).

[4]  Mike Thelwall,et al.  Sentiment in short strength detection informal text , 2010 .

[5]  Jia Rong,et al.  Deep learning for pollen allergy surveillance from twitter in Australia , 2019, BMC Medical Informatics and Decision Making.

[6]  Lahari Poddar,et al.  Modeling Temporal Progression of Emotional Status in Mental Health Forum: A Recurrent Neural Net Approach , 2017, WASSA@EMNLP.

[7]  Mike Thelwall,et al.  The Heart and Soul of the Web? Sentiment Strength Detection in the Social Web with SentiStrength , 2017 .

[8]  Ali A. Minai,et al.  Semantic knowledge inference from online news media using an LDA-NLP approach , 2011, The 2011 International Joint Conference on Neural Networks.

[9]  S. Strathdee,et al.  The coronavirus 2019-nCoV epidemic: Is hindsight 20/20? , 2020, EClinicalMedicine.

[10]  Tara N. Sainath,et al.  Convolutional, Long Short-Term Memory, fully connected Deep Neural Networks , 2015, 2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[11]  Basemah Alshemali,et al.  Improving the Reliability of Deep Neural Networks in NLP: A Review , 2020, Knowl. Based Syst..

[12]  D. Heron,et al.  Reddit and Radiation Therapy: A Descriptive Analysis of Posts and Comments Over 7 Years by Patients and Health Care Professionals , 2019, Advances in radiation oncology.

[13]  Gonzalo A. Ruz,et al.  Sentiment analysis of Twitter data during critical events through Bayesian networks classifiers , 2020, Future Gener. Comput. Syst..

[14]  Balamurugan Ramasamy,et al.  Classification of healthcare data using hybridised fuzzy and convolutional neural network , 2019, Healthcare technology letters.

[15]  Hardik Meisheri,et al.  Sentiment Extraction from Consumer-Generated Noisy Short Texts , 2017, 2017 IEEE International Conference on Data Mining Workshops (ICDMW).

[16]  Mike Conway,et al.  Tracking Health Related Discussions on Reddit for Public Health Applications , 2017, AMIA.

[17]  Christopher S. G. Khoo,et al.  Sentiment lexicons for health-related opinion mining , 2012, IHI '12.

[18]  Javier Palanca,et al.  Semantic-based padding in convolutional neural networks for improving the performance in natural language processing. A case of study in sentiment analysis , 2020, Neurocomputing.

[19]  Kyung-shik Shin,et al.  Deep learning models and datasets for aspect term sentiment classification: Implementing holistic recurrent attention on target-dependent memories , 2020, Knowl. Based Syst..

[20]  Jian Ma,et al.  Sentiment classification: The contribution of ensemble learning , 2014, Decis. Support Syst..

[21]  Jingyu Huang,et al.  Optimization of Recurrent Neural Networks on Natural Language Processing , 2019, ICCPR.

[22]  Om Prakash Sangwan,et al.  Deep Learning Approaches for Textual Sentiment Analysis , 2020 .

[23]  He He,et al.  GluonCV and GluonNLP: Deep Learning in Computer Vision and Natural Language Processing , 2020, J. Mach. Learn. Res..

[24]  Nicolae Goga,et al.  [WiP] Sentiment Analysis Electronic Healthcare System Based on Heart Rate Monitoring Smart Bracelet , 2018, 2018 IEEE 11th Conference on Service-Oriented Computing and Applications (SOCA).

[25]  Adil Rajput,et al.  Natural Language Processing, Sentiment Analysis and Clinical Analytics , 2019, Innovation in Health Informatics.

[26]  Massimo Piccardi,et al.  Recurrent neural networks with specialized word embeddings for health-domain named-entity recognition , 2017, J. Biomed. Informatics.

[27]  Qingcai Chen,et al.  Fuzzy deep belief networks for semi-supervised sentiment classification , 2014, Neurocomputing.

[28]  Michael I. Jordan,et al.  Latent Dirichlet Allocation , 2001, J. Mach. Learn. Res..

[29]  Joel H. Saltz,et al.  Social Media Based Analysis of Opioid Epidemic Using Reddit , 2018, AMIA.

[30]  Hamid Naderi,et al.  Similarity of medical concepts in question and answering of health communities , 2020, Health Informatics J..

[31]  Xia Feng,et al.  Latent Dirichlet allocation (LDA) and topic modeling: models, applications, a survey , 2017, Multimedia Tools and Applications.

[32]  Andrew McCallum,et al.  Gibbs Sampling for Logistic Normal Topic Models with Graph-Based Priors , 2008 .

[33]  Christopher C. Yang,et al.  Enriching User Experience in Online Health Communities Through Thread Recommendations and Heterogeneous Information Network Mining , 2018, IEEE Transactions on Computational Social Systems.

[34]  ThelwallMike,et al.  Sentiment strength detection in short informal text , 2010 .

[35]  Balamurali Ananthanarayanan,et al.  Develop a Neural Model to Score Bigram of Words Using Bag-of-Words Model for Sentiment Analysis , 2020 .

[36]  Melissa Roy,et al.  Ebola and Localized Blame on Social Media: Analysis of Twitter and Facebook Conversations During the 2014–2015 Ebola Epidemic , 2019, Culture, Medicine, and Psychiatry.

[37]  Alhareth Mohammed Abu Hussein,et al.  Sentiment Analysis in Healthcare: A Brief Review , 2019 .

[38]  Paul Buitelaar,et al.  Unsupervised Classification of Health Content on Reddit , 2019, PDH.

[39]  Kristina Lerman,et al.  Modeling Social Annotation: A Bayesian Approach , 2008, TKDD.

[40]  Juntae Yoon,et al.  Link-topic model for biomedical abbreviation disambiguation , 2015, J. Biomed. Informatics.

[41]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[42]  V. G. Vinod Vydiswaran,et al.  Identifying peer experts in online health forums , 2019, BMC Medical Informatics and Decision Making.

[43]  Keun Ho Ryu,et al.  Ontology-Based Healthcare Named Entity Recognition from Twitter Messages Using a Recurrent Neural Network Approach , 2019, International journal of environmental research and public health.

[44]  Jina Huh,et al.  Detecting clinically related content in online patient posts , 2017, J. Biomed. Informatics.

[45]  Hongfang Liu,et al.  Assessing Unmet Information Needs of Breast Cancer Survivors: Exploratory Study of Online Health Forums Using Text Classification and Retrieval , 2018, JMIR cancer.

[46]  Yang Zhang,et al.  Combining paper cooperative network and topic model for expert topic analysis and extraction , 2017, Neurocomputing.

[47]  Yuan Luo,et al.  Recurrent Neural Networks for Classifying Relations in Clinical Notes , 2017, AMIA.

[48]  Mike Thelwall,et al.  Topic-based sentiment analysis for the social web: The role of mood and issue-related words , 2013, J. Assoc. Inf. Sci. Technol..