Discovering social events through online attention

Area of Science:

• Computational Social Science
• Economics
• Social Media Analysis

Background:

• Twitter is a major social media platform with high daily traffic.
• Identifying crisis events amidst heavy social media data is challenging.
• Social media is increasingly used for organizing and relief during crises.

Purpose of the Study:

• To develop a new approach for detecting crisis events on Twitter.
• To assess the effectiveness of different "thermostats" for social media activity analysis.
• To combine computational social science and economic methods for improved crisis detection.

Main Methods:

• Assessed various "thermostats" of social media activity.
• Compared existing literature methods with an economic method.
• Integrated computational social science and economic approaches.

Main Results:

• Developed an effective approach to locate crisis events in large Twitter datasets.
• Demonstrated the method's strength in identifying significant social events.
• Successfully located social events related to the Occupy Wall Street movement.

Conclusions:

• The combined approach effectively identifies crisis periods on Twitter.
• This method aids in distinguishing turmoil from typical discussion.
• Applicable to analyzing social events and crisis communication on social media.