4 August 2021 in Cluster Computing
18 January 2021 in Peer-to-Peer Networking and Applications
3 April 2017 in American Journal of Distance Education
The open nature of Massive Open Online Courses (MOOCs) attracts a large number of learners with different backgrounds, skills, motivations, and goals. This has brought a need to understand such heterogeneity in populations of MOOC learners. Categorizing these learners based upon their interaction with the course can help address this need and suggest possible improvements in course design and delivery. In this article, the K-means clustering technique with careful seeding is used to obtain clusters of learners having similar interaction in the course. Learners are grouped based on their interaction with course material, video lectures, discussion forums, and assessments. In the analysis of thirteen courses, the proposed method identified learners’ classes as Uninterested, Casuals, Performers, Explorers and Achievers. Each class of learners had distinct interaction with the course and followed a certain learning approach. The learners’ classes were mapped to the standard surface, deep, and strategic learning approaches.This article also highlights the data preparation phase and its importance in data mining.
15 September 2022 in Computers and Electrical Engineering
7 March 2020 in International Journal of System Assurance Engineering and Management
18 April 2024 in Blockchain in Healthcare Today
Blockchain technology has emerged as a pivotal point to enhance privacy and security in enterprise applications and cyber world. However, scalability is an issue researcher are grappling with, in large enterprises, especially in organizations bearing multiple levels of hierarchy and access privilege. Therefore, the existing models and consensus algorithms suffer one way or another. The medical or healthcare sector suffers this problem the most due to the huge amount of data and probably the central point of failure of the traditional database management system. This paper addresses the situation through a hierarchical model in Hyperledger fabric enterprise application through a healthcare sector use case. Multiple organizations are added to each hierarchy considering them as different organization levels (Hospitals, Hospital Governance, and Insurance company). Currently the first implementation has two levels of hierarchy to show networks of hospitals joining an Insurance Company. Our primary experiment revolves around this model to test and enhance the performance of the network. Performance of the model is assessed by varying and scaling environmental parameters such as the number of organizations, transaction numbers, channels, block intervals and block sizes. The benchmarking tool used is Hyperledger caliper to test various indicators such as success and failure rates along with throughput and latency. The current work only tests the scalability of the model with patient data.