Name
Title | ||
|---|---|---|
Understanding Popularity Evolution Patterns Of Hot Topics Based On Time Series Features |
Abstract | ||
|---|---|---|
Understanding popularity evolution patterns of hot topics is important for online recommendation systems and marketing services. Previous research has analyzed popularity evolution patterns based on the time series which only captures the feature of peaks. However, hot topics experience more complex popularity evolution patterns, not only peaks but also other time series features: level, trend and seasonality. In this paper, we present a method to model and understand popularity evolution patterns based on the three time series features for two types of hot topics: burst and non-burst. Our experimental results demonstrate that the seasonality of the time series is multiplicative, which means the size of the fluctuations in popularity evolution pattern of a hot topic varies with the change of trend. The level and trend of the time series are relatively unstable for burst topics compared with non-burst topics. |
Year | DOI | Venue |
|---|---|---|
2014 | 10.1007/978-3-319-11119-3_6 | WEB TECHNOLOGIES AND APPLICATIONS, APWEB 2014, PT II |
Keywords | Field | DocType |
Popularity Evolution Patterns, Hot Topics, Time Series | Recommender system,Data science,Data mining,Computer science,Popularity,Seasonality | Conference |
Volume | ISSN | Citations |
8710 | 0302-9743 | 2 |
PageRank | References | Authors |
0.37 | 16 | 5 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Changjun Hu | 1 | 130 | 27.56 |
| Ying Hu | 2 | 13 | 2.30 |
| Wenwen Xu | 3 | 8 | 3.17 |
| Peng Shi | 4 | 5 | 2.18 |
| Shushen Fu | 5 | 12 | 1.28 |