New Trends in Over the Top Media Service (OTT) Web User Behaviour Analysis and Unethical User Prediction

被引:2
作者
Nguyen, Ha Huy Cuong [1 ]
Grzonka, Daniel [2 ]
Khoa, Bui Thanh [3 ]
Sagar, K. V. Daya [4 ]
Abbasi, Irshad Ahmed [5 ]
Mahaveerakannan, R. [6 ]
Alkhayyat, Ahmed [7 ]
机构
[1] Univ Danang, Da Nang, Vietnam
[2] Cracow Univ Technol, Krakow, Poland
[3] Ind Univ Ho Chi Minh City, Ho Chi Minh, Vietnam
[4] Koneru Lakshmaiah Educ Fdn, Dept Elect & Commun Engn, Vaddeswaram, Andhra Prades, India
[5] Belqarn Univ Bisha, Fac Sci & Arts, Dept Comp Sci, Sabt Al Alaya 61985, Saudi Arabia
[6] Saveetha Inst Med & Tech Sci, Saveetha Sch Engn, Dept Comp Sci & Engn, Chennai, Tamil Nadu, India
[7] Islamic Univ, Coll Tech Engn, Najaf, Iraq
关键词
Bitrate-adaptivity; Content delivery network; Network functions virtualization; Over-the-top media service; Quality of experience and user prediction; OVER-THE-TOP; STREAMING SERVICES;
D O I
10.1007/s11036-023-02214-7
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
This research shows unethical user prediction on the peer-to-peer content delivery network against the Over-The-Top media service (OTT). The Web User Behaviour Analysis and unethical user prediction are performed based on the OTT multimedia service platform. The Analysis and unethical user prediction contain the four stages of processing. Firstly "Peer to Peer Content Delivery Network" is used to provide services and ideal resources (User). It uses ResourceCache: an optimized algorithm for giving efficient CDN-based over-the-top video streaming services. Secondly, "Selective destination Bitrate-adaptivity" is employed to provide the OTT Service (Devices). It utilizes bitrate adaptivity to improve the adaptivity of Over-The-Top Television systems. Thirdly, the "Network Functions Virtualization (NFV)" enabled multi-access edge data centers to embed to maintain sufficient data about users and their respective devices. It uses the QoE-based Load Balancing of OTT video content in CDN networks to improve the QoE of OTT multimedia services in wireless scenarios. Finally, the efficient "Review of Recommender System" are taken to collect feedback about services and subscription. It is most useful to check unethical user access. The performance result of the proposed system conceives the analysis based on web users' behavior prediction accuracy of 76%, Average precision of 67.8%, Average recall of 74.2%, and Average f1-measures of 85.9% against the performance of unethical user prediction.
引用
收藏
页码:1811 / 1827
页数:17
相关论文
empty
未找到相关数据