Mpox Narrative on Instagram: A Labeled Multilingual Dataset of Instagram Posts on Mpox for Sentiment, Hate Speech, and Anxiety Analysis

Overview

• The paper presents a multilingual dataset of 60,127 Instagram posts about the mpox (monkeypox) outbreak, which occurred between July 23, 2022, and September 5, 2024.
• The dataset, available at https://dx.doi.org/10.21227/7fvc-y093, contains posts in 52 languages.
• For each post, the dataset includes attributes such as Post ID, Post Description, Date of publication, Language, and an English translation.
• The authors also performed sentiment analysis, hate speech detection, and anxiety/stress detection on the posts and included those results as separate attributes.

Plain English Explanation

This research paper focuses on social media mining during the recent mpox outbreak. The researchers developed a comprehensive dataset of over 60,000 Instagram posts related to mpox, which were published between July 2022 and September 2024.

The dataset includes posts in 52 different languages, and for each post, the researchers provide key information like the post's ID, text, publication date, and language. Importantly, they also translated each post into English using a machine translation API.

After compiling the dataset, the researchers performed several analyses on the posts. They used sentiment analysis to categorize the emotional tone of each post, such as fear, joy, or anger. They also checked for the presence of hate speech and signs of anxiety or stress.

This comprehensive dataset and the accompanying analyses provide valuable insights into how people were discussing and reacting to the mpox outbreak on social media. Researchers and public health officials could use this information to better understand the public's concerns and responses during disease outbreaks.

Technical Explanation

The researchers first collected a dataset of 60,127 Instagram posts about the mpox outbreak, which occurred between July 23, 2022, and September 5, 2024. These posts were written in 52 different languages.

For each post, the dataset includes the following attributes:

• Post ID
• Post Description
• Date of publication
• Language
• English translation (using Google Translate API)

After compiling the dataset, the researchers performed several analyses on the posts:

1. Sentiment Analysis: They classified each post into one of seven sentiment classes: fear, surprise, joy, sadness, anger, disgust, or neutral.
2. Hate Speech Detection: They identified whether each post contained hate speech or not.
3. Anxiety/Stress Detection: They determined whether each post indicated any signs of anxiety or stress.

The results of these analyses were included as additional attributes in the dataset.

The sentiment analysis revealed that 27.95% of the posts expressed fear, 2.57% expressed surprise, 8.69% expressed joy, 5.94% expressed sadness, 2.69% expressed anger, 1.53% expressed disgust, and 50.64% were neutral.

In terms of hate speech, 95.75% of the posts did not contain any hate speech, while 4.25% did.

Finally, the researchers found that 72.05% of the posts did not indicate any signs of anxiety or stress, while 27.95% did.

Critical Analysis

The researchers have provided a valuable dataset and analysis that can help researchers and public health officials better understand how people were discussing and reacting to the mpox outbreak on social media.

One potential limitation of the research is that the dataset is limited to Instagram posts, which may not represent the full breadth of online conversations about mpox. It would be useful to expand the analysis to other social media platforms, such as Twitter or Reddit, to get a more comprehensive understanding of the public's response.

Additionally, the researchers did not provide much information about the accuracy or reliability of the sentiment analysis, hate speech detection, and anxiety/stress detection methods used. It would be helpful to have more details on the performance of these algorithms, as well as any potential biases or limitations.

Despite these minor concerns, the dataset and analyses presented in this paper represent an important contribution to the field of social media mining and disease outbreak research. Researchers and public health officials can use this information to better understand the public's concerns and reactions during future disease outbreaks, potentially informing their communication and response strategies.

Conclusion

This research paper presents a comprehensive dataset of over 60,000 Instagram posts about the mpox outbreak, along with detailed analyses of the sentiment, hate speech, and anxiety/stress expressed in those posts.

The dataset, which is publicly available, provides valuable insights into how people were discussing and reacting to the mpox outbreak on social media. The researchers' findings suggest that a significant portion of the posts expressed fear, as well as some level of anxiety or stress, while a small percentage contained hate speech.

This research can be used by researchers and public health officials to better understand the public's response to disease outbreaks and develop more effective communication strategies. By leveraging social media data and analysis, they can gain valuable insights into the concerns and sentiments of the general public, which can inform their public health policies and interventions.