Net-TF-SW: Event Popularity Quantification with Network Structure.

Area of Science:

• Social media analysis
• Information science
• Crisis informatics

Background:

• Event popularity analysis is crucial for understanding online trends and managing information during crises.
• Effective crisis communication requires accurate tracking of event popularity to prevent the spread of misinformation.

Purpose of the Study:

• To propose Net-TF-SW, a novel method for noise-robust and explainable topic popularity analysis.
• To evaluate the method's effectiveness using real-world crisis data from social media.

Main Methods:

• Net-TF-SW (Network Topic Frequency - Sliding Window) method for topic popularity quantification.
• Application to Twitter data concerning the COVID-19 pandemic and the Fukushima Daiichi Nuclear Disaster.
• Comparative analysis against existing event popularity analysis methods.

Main Results:

• The proposed Net-TF-SW method demonstrates superior robustness against noise compared to existing techniques.
• Accurate quantification of event popularity was achieved for significant crisis events.
• The method proved effective in analyzing trends during major public anxiety events.

Conclusions:

• Net-TF-SW offers a reliable approach for analyzing event popularity, particularly in noisy social media environments during crises.
• The method aids in understanding public trends and facilitating appropriate information dissemination.
• This research contributes to more effective crisis informatics and misinformation management.