Abstract
Negative emotional contagion is spreading in social networks and is adversely affecting people; it can even lead to depression and suicide. By implementing the genetic algorithm along with community detection algorithms, this article aims to uncover the Twitter features that enhance the spread of negative emotions on the network. The novelty of this article is that it focuses on a combination of community detection features that enhance the diffusion of negative emotions in social networks. The genetic algorithm benefits the study as it uses the modularity of the community detection algorithms (Louvain and Label Propagation algorithms) as fitness values in order to find the most favorable values in cost-effective manner. While other studies have concentrated on singular Twitter features, applying the Louvain genetic algorithm and the Label Propagation genetic algorithm to negative emotion data from Twitter resulted in higher modularity from a combination of features than the results from single features. This demonstrates that a combination of features increases the diffusion of negative emotions in Twitter communities.