Algorithm 1.

Frequent Temporal Pattern Mining Algorithm

1:	procedure frequent_temporal_pattern_mining($DS,{min}_{\mathit{freq}}$)
2:	Initialize $D{S}^{*}$	▷ An empty set to record $CIG\mathrm{s}$ associated with temporal networks.
3:	for $T\in DS$ do
4:	Initialize $CIG$ with empty vertex and edge sets
5:	for $edge\in T$ do
6:	Add a vertex $v$ to $CIG$ corresponding to $edge$
7:	Label $v$ in $CIG$ with the endpoints’ and edge’s labels and duration of $edge$
8:	Identify other overlapping edges $e^{\prime }\mathrm{s}$ in $T$ read prior to $edge$ and with vertices in common with $edge$ and with vertices in common with $edge$
9:	Connect $e^{\prime }\mathrm{s}$ corresponding vertices in $CIG$ to $v$ with directed edges
10:	Label the connections with their starting points’ differences in $T$
11:	Append $CIG$ to $D{S}^{*}$
12:	frequent_temporal_patterns = PATTERN_MINING($D{S}^{*},{min}_{\mathit{freq}}$)
13:	Return frequent_temporal_patterns