From: Estimating tie strength in social networks using temporal communication data
Type | Variable | Name | Cluster | Description |
---|---|---|---|---|
I | w | Number of calls/contacts | \(C_{1}\) | Late night and early morning |
I | l | Total call duration | \(C_{2}\) | Monday early morning |
I | l̄ | Average call duration | \(C_{3}\) | Monday early morning |
I | r | Reciprocity | \(C_{4}\) | Weekday 7 am |
AP | \(a_{d}\) | Active days | \(C_{5}\) | Weekday afternoon |
AP | \(a_{h}\) | Active hours | \(C_{6}\) | Weekday evening |
IET | τ̄ | Mean IET | \(C_{7}\) | Weekday early morning |
IET | \(\sigma _{\tau }\) | Std. Dev. of IET | \(C_{8}\) | Thursday early morning |
IET | B | Burstiness Coefficient | \(C_{9}\) | Weekend evening |
IET | \(\bar{\tau }_{R}\) | Average Relay Time | \(C_{10}\) | Weekend morning |
TS | f̂ | Relative freshness | \(C_{11}\) | Saturday Morning |
TS | age | Age | \(C_{12}\) | Weekend afternoon |
TS | TS | Temporal Stability | \(C_{13}\) | Saturday late afternoon |
BC | \(N^{E}\) | Number of busty events | \(C_{14}\) | Sunday morning |
BC | Ē | Average calls per bursty event | \(C_{15}\) | Sunday afternoon |
BC | \(\sigma ^{E}\) | Std. Dev. of event distribution | \(C^{*}\) | Vector of clusters |
BC | \(CV^{E}\) | CV of event distribution | ||
DBC | t̄ | Avg. interaction time | ||
DBC | \(\sigma _{t}\) | Std. Dev. of interaction times | ||
DBC | log(T) | Test statistic for t̄ | ||
DP | JSD | Differences in daily behaviour |