Towards data mining in IoT cloud computing networks : Collaborative filtering based recommended system. (4th July 2021)

• Record Type: Journal Article
• Authors: Kumar, Mukesh; Kumar, Sushil; Kashyap, Panjak Kumar
• Is Part Of: Journal of discrete mathematical sciences & cryptography. Volume 24:Number 5(2021)
• Page Start: 1309
• Page End: 1326
• Publication Date: 2021-07-04
• Subjects: 68 (Computer Science); Internet of Things -- Collaborative filtering -- Data mining -- Time weight decay function -- Accuracy rate; Computer science -- Mathematics -- Periodicals; Cryptography -- Periodicals; Computer science -- Mathematics; Cryptography; Periodicals; 004.0151
• DOI: 10.1080/09720529.2021.1932918
• Languages: English
• ISSNs: 0972-0529

Abstract

Abstract: The ever-increasing data due to gaining popularity of Internet of Things (IoT) needs to be adequate storage to compute complex task with accuracy and low latency and finally recommend the user's preference in the near future. The cloud computing and data mining technique having ability to provide open platform for communication and generate precise recommendation. In this regard, this paper mainly carried out two aspects: firstly, we built four layers IoT cloud computing architecture that provides an open platform for communication with various heterogeneous multi-source things. Secondly, we present a recommended system model based on collaborative filtering algorithm to enhance the accuracy rate of the items in the top priority recommended list. The proposed model inherently utilizes the user-item's scoring matrix, asymmetrical influence degree on the similar items between users and time weight decay function for the user's preferences. Finally, extensive simulations are done to show the accuracy rate, loss rate and recall rate of recommendation for the proposed model. Further, comparative analysis of results proved that our proposed model outperforms the other state-of-art model in terms of accuracy rate of the recommendation on the item with respect to data sample set size.