Name
Abstract | ||
|---|---|---|
This paper presents a supervised no-reference model for QoE assessment on IPTV services. MOS_V (Mean Opinion Score for Video) which has five levels is recommended as subjective assessment indicator, while MDI (media delivery index), PCR (program clock reference) as well as MR (media rate) are recommended as objective assessment indicators. Collecting and using both of them to build the prediction model is remarkable. The basic decision classification trees in C4.5 algorithm and Random Forests are chosen to build the model. Based on them, two innovative measures are proposed to improve model accuracy. Considering uneven rating levels of people, an adaptive rating modification algorithm is put forward for MOS_V to adjust all rating levels to the same value coordinate system. Besides, owing to coarse granularity of traditional MOS_V, we adjust classification width and fault tolerance rate correspondingly. The experimental results show that the model outperforms traditional methods which only apply machine learning methods in prediction accuracy. |
Year | DOI | Venue |
|---|---|---|
2016 | 10.1109/CCIS.2016.7790268 | 2016 4th International Conference on Cloud Computing and Intelligence Systems (CCIS) |
Keywords | Field | DocType |
QoE Assessment,MOS_V,IPTV,Adaptive rating modification,Classification width adjustment | Data mining,Computer science,Media Delivery Index,Mean opinion score,Fault tolerance,Program clock reference,Granularity,IPTV,Random forest | Conference |
ISSN | ISBN | Citations |
2376-5933 | 978-1-5090-1257-2 | 0 |
PageRank | References | Authors |
0.34 | 6 | 3 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Quangu Chen | 1 | 0 | 0.34 |
| Yuehui Jin | 2 | 31 | 9.06 |
| Tan Yang | 3 | 23 | 10.97 |