Inspecting each link and then computing the modularity difference requires $O(L^{**}(K))$ computations,

where $L$ is the number of links and $K$ is the number of operations to calculate the modularity

difference.

The algorithm always performs $N$ community mergings.

For a dense graph the run time complexity is $O(N^{{?}^{{?}^{?}}}2).$

With appropriate data$-$structures the run time complexity can be reduced to $O(NlogN)$Which of the following statements about the complexity of the greedy modularity maximization algorithm is TRUE :