In this chapter, we focus on the multimedia service data preprocessing and feature extraction, which are core steps in the multimedia QoE evaluation. Specifically, multimedia service data collection and preprocessing is firstly introduced. We briefly describe three representative datasets: (1) IPTV service dataset collected by operator; (2) OTT service dataset collected from students in Nanjing University of Posts and Telecommunications; (3) dataset crawled across the web. Then, we highlight our works about feature extractions for user-related IFs, containing five aspects: viewing time ratio calculation, user interest inference, user type classification, user behavior analysis, and user emotion parsing from their comments in danmaku.
This is a preview of subscription content, log in to check access.
Huang R, Wei X, Gao Y, Lv C, Mao J, Bao Q (2018) Data-driven QoE prediction for IPTV service. Comput Commun 118(12):195–204CrossRefGoogle Scholar
Gao Y, Wei X, Zhang X, Zhuang W (2018) A combinational LDA-based topic model for user interest inference of energy efficient IPTV service in smart building. IEEE Access 6:48921–48933CrossRefGoogle Scholar
Zhang Y, Chen W, Zha H, Gu X (2015) A time-topic coupled LDA model for IPTV user behaviors. IEEE Trans Broadcast 61(1):51–65CrossRefGoogle Scholar
Kate RJ (2016) Using dynamic time warping distances as features for improved time series classification. Data Min Knowl Disc 30(2):283–312MathSciNetCrossRefGoogle Scholar
Hao Y, Yang J, Chen M (2019) Emotion-aware video QoE assessment via transfer learning. IEEE Multimed 26(1):31–40CrossRefGoogle Scholar
Chen Y, Gao Q, Rau P-LP (2017) Watching a movie alone yet together: understanding reasons for watching danmaku videos. Int J Hum Comput Interact 33(9):731–743CrossRefGoogle Scholar
Huang R (2018) Research on the theory and application of the key technologies of user quality of experience video service. Doctor Thesis, Nanjing University of Posts and Telecommunications (Supervisor: Zhu H, Zhou L)Google Scholar