Abstract
This paper focuses on some principal pre- and post-events-related to the 2018 DR Congo general elections to identify controversial and divisive topics that sparked political tensions and awakened long-simmering tribal cleavages within the Congolese community. Thus, we detect social communities yielded by these events and examine their structural features and the content of interactions. We measure divergence of opinion within different social communities and then identify intransigent and compliant social communities. Furthermore, we discover affinity relationships between members of different social communities for predicting (i) nascent uncompromising social communities and (ii) whether adamant social communities are likely to change their opinions and minds over time. Extensive experimental comparisons with several best-performing community detection-type techniques have confirmed the effectiveness of our method and demonstrated that this latter provides competitive results. We found that 64.7% of social communities nurture fierce rivalry between them and 1.9% of social communities assemble members with neutral opinions.