Online election campaigning: Identifying political parties using likes and comments

• Published in: Complex Networks & Their Applications V Vol. 693; pp. 619 - 630
• Main Authors: Liebig, Jessica, Khairuddin, Mohammad Adib, Rao, Asha
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer International Publishing AG 01.01.2017; Springer International Publishing
• Series: Studies in Computational Intelligence
• Subjects: Bipartite Network; Community Detection Algorithm; Election Campaign; National Front; Opposing Party
• ISBN: 9783319509006; 3319509004
• ISSN: 1860-949X; 1860-9503
• DOI: 10.1007/978-3-319-50901-3_49

Politicians use social media to engage and communicate with voters, in particular during election campaigns. This article investigates data collected from politicians’ Facebook pages during the 2013 Australian Federal election and the 2013 Malaysian General election. We wish to gain insight into whether the likes and comments of Facebook users reflect actual connections between politicians during an election campaign. Intuitively, a Facebook user who supports a particular party would not like the posts published by candidates who are associated with opposing parties. However, we observe that users often like the posts by candidates belonging to opposing parties. Our analysis of the data shows that many of the likes and comments made by Facebook users are statistically insignificant. Deletion of these insignificant likes and comments clearly reveals the different parties of the political system. In this paper we consider only the topology of the network representing the datasets, presenting an alternative to the often cumbersome sentiment analysis.