Effective Semisupervised Community Detection Using Negative Information

<div>Results (NMI) of the proposed algorithm with different selections of NL and TL in the LFR benchmark networks. Each point in the graph represents the average over 100 runs on randomly generated LFR benchmark networks with the given parameters.</div>

Mathematical Problems in Engineering

fig4

Figure 4

Figure 4: Effective Semisupervised Community Detection Using Negative Information 