Evolutionary Prediction of Nonstationary Event Popularity Dynamics of Weibo Social Network Using Time-Series Characteristics

Figure 6

Stationary sequence recognition result on the event Chengdu-Driver: (a) forwarding longitudinal series 1–4; (b) forwarding longitudinal series 3–4. The sum of the forwarding ratios with volatility values greater than 1 equals 0.78 which is shown in (a). The score indicates the nonstationary characteristic of the event. The instability is mainly caused by the forwarding behavior of levels 3 and 4 according to the value 0.88 shown in (b).

(a)

(b)