Exploring the Dynamics of Targeting and Non-Targeting Rumour Spreaders on Twitter: A Qualitative Analysis
Name
Yu-Chen Yang
Abstract
Misinformation has gained much notoriety for severe damage to our daily life. With the accessibility of the internet and the popularity of social media, misinformation has become even more challenging than ever. Many people fall into the targets of misinformation, and social cohesion is hence threatened. Although researchers have been devoted to studying targeting misinformation, there is a lack of studies on targeting rumour. However, it is crucial to understand targeting rumours for rumour detection. In this thesis, we identified the targeted groups in the PHEME-9 rumour dataset and annotated the targeting text. The targeted groups in this work were an ethnic group of people, such as the black and Muslims, being blamed for an incident. To aid in the identification of characteristics of targeting tweets, we have carried out an exploratory analysis of targeting tweets into three dimensions. From topic analysis, targeting tweets were mainly concerned with social issues, while non-targeting tweets discussed various issues. Additionally, we found that negative targeting tweets stirred up extremely negative in their following non-targeting tweets. Furthermore, we looked into the targeting users and provided a possible way of reducing targeting users. In summary, this thesis contributed to the linguistic characteristics of targeting tweets and provided possible actions to combat targeting rumours. We also extended the existing rumour dataset with targeting labels.
Graduation Thesis language
English
Graduation Thesis type
Master - Innovation and Technology Management
Supervisor(s)
Shakshi Sharma, Rajesh Sharma
Defence year
2023