A Study of the Relationship Between the Geographic Locations of the User and Participation in Twitter During Different Types of News Events

Abstract

Twitter is one of the most active social networks in news sharing. People report local events sometimes faster than news agencies. During major events such as earthquakes and presidential elections, people share tweets, retweets, images and links related to the event, creating an overwhelming number of posts. The amount of data generated by social media provides a powerful tool for knowledge discovery for individuals, organizations and governments. However, the large stream of tweets makes tracking interesting posts a challenging task. Understanding user behavior during different events provides substantial knowledge to get the most out of the social media power. A twitter post encapsulates several fields about the tweet and the users, such as posting date, users locations, time zones, and geographic coordinates. The main aim of this work is to investigate the relationship between user participation, news type and geographic locations of users in Twitter. To achieve this goal, posts related to a certain event were retrieved by keyword search, and geographic information was obtained from users’ profiles, then posts were analyzed by news type and geographic location. The results showed that finance news tweets had distinct user behavior, finance tweets have more original tweets, more links and less hashtags, while politics tweets had the largest number of hashtags, compared to disaster and finance tweets. The investigation of the relationship between users’ participation and news type would provide guidelines for social media management and advertisement, similarly the analysis of the relationship between the participation and users’ locations provides insight for information diffusion research and location-aware news recommendation systems.