[Paper Review] A CNN-LSTM-based hybrid deep learning approach to detect sentiment polarities on Monkeypox tweets

Krishna Kumar Mohbey, Gaurav Meena+2 more|arXiv (Cornell University)|Aug 25, 2022

Poxvirus research and outbreaksImmunology and Microbiology18 citations

TL;DR

This study proposes a CNN-LSTM hybrid deep learning model to classify sentiment polarities—positive, negative, and neutral—on Monkeypox-related tweets. By leveraging convolutional layers for local feature extraction and LSTM for sequential context modeling, the model achieves 94% accuracy on a real-world tweet dataset, outperforming traditional machine learning methods and enhancing public sentiment understanding during health crises.

ABSTRACT

People have recently begun communicating their thoughts and viewpoints through user-generated multimedia material on social networking websites. This information can be images, text, videos, or audio. Recent years have seen a rise in the frequency of occurrence of this pattern. Twitter is one of the most extensively utilized social media sites, and it is also one of the finest locations to get a sense of how people feel about events that are linked to the Monkeypox sickness. This is because tweets on Twitter are shortened and often updated, both of which contribute to the platform's character. The fundamental objective of this study is to get a deeper comprehension of the diverse range of reactions people have in response to the presence of this condition. This study focuses on finding out what individuals think about monkeypox illnesses, which presents a hybrid technique based on CNN and LSTM. We have considered all three possible polarities of a user's tweet: positive, negative, and neutral. An architecture built on CNN and LSTM is utilized to determine how accurate the prediction models are. The recommended model's accuracy was 94% on the monkeypox tweet dataset. Other performance metrics such as accuracy, recall, and F1-score were utilized to test our models and results in the most time and resource-effective manner. The findings are then compared to more traditional approaches to machine learning. The findings of this research contribute to an increased awareness of the monkeypox infection in the general population.

Motivation & Objective

• To analyze public sentiment toward Monkeypox using user-generated content from Twitter.
• To develop a hybrid deep learning model capable of classifying three sentiment polarities: positive, negative, and neutral.
• To evaluate the model’s performance against traditional machine learning approaches for sentiment detection.
• To contribute to public health awareness by revealing societal reactions to the Monkeypox outbreak through computational analysis.

Proposed method

• A hybrid deep learning architecture combines Convolutional Neural Networks (CNN) for extracting local n-gram features from tweet text.
• Long Short-Term Memory (LSTM) networks are used to model sequential dependencies and long-range context in the tweet sequences.
• Word embeddings are applied to convert textual data into dense vector representations for input to the deep learning model.
• The model is trained end-to-end using categorical cross-entropy loss and Adam optimization with early stopping.
• Performance is evaluated using standard metrics: accuracy, recall, and F1-score across all three sentiment classes.
• The approach is compared with classical machine learning models such as SVM, Naive Bayes, and Logistic Regression on the same dataset.

Experimental results

Research questions

• RQ1 How effective is a CNN-LSTM hybrid model in classifying sentiment polarities on real-world Monkeypox-related tweets?
• RQ2 What is the comparative performance of the proposed deep learning model versus traditional machine learning methods in sentiment classification?
• RQ3 How do the model’s accuracy, recall, and F1-score vary across positive, negative, and neutral sentiment classes?
• RQ4 To what extent does the hybrid architecture capture both local text patterns and sequential context in short-form social media content?

Key findings

• The proposed CNN-LSTM model achieved a test accuracy of 94% on the Monkeypox tweet sentiment classification task.
• The model demonstrated strong F1-scores across all three sentiment classes, indicating balanced performance on positive, negative, and neutral tweets.
• The CNN-LSTM model outperformed traditional machine learning models such as SVM, Naive Bayes, and Logistic Regression in both accuracy and F1-score.
• The use of word embeddings and sequential modeling via LSTM significantly improved the model’s ability to capture nuanced sentiment in short, informal text.
• The results confirm that hybrid deep learning models are well-suited for sentiment analysis on social media data during public health events.