On the Evaluation of Community Detection Algorithms on Heterogeneous Social Media Data


One fundamental problem in social networks is the identification of groups of elements (also known as communities) when group membership is not explicitly available. Community detection has proven to be valuable in diverse domains such as biology, social sciences and bibliometrics. Thus, several community detection techniques have been developed. Nonetheless, as real networks are very heterogenous, the question of how communities should be assessed remains open. Whilst there are several works that have analysed the performance of diverse community detection algorithms over artificial graph benchmarks, the evaluation over real social networks has received comparatively less attention. Motivated by the lack of such studies, this chapter focuses on the analysis of the performance of community detection algorithms over social media networks, and the quantification of the structural properties of the discovered communities.

Linking and Mining Heterogeneous and Multi-view Data, pp. 295–333, https://doi.org/10.1007/978-3-030-01872-6_12