Dynamics of Knowledge Management in 4IR Through HR Interventions: Conceptual Framework

Resource allocation for big data streams in cloud systems involves select-
ing the appropriate cloud resources. Big data has certain precise features such as size,

speed, veracity, variety, and value. In this paper, a workload forecasting system for

resource allocation in big data streams is developed. In this system, the data char-
acteristics such as data type (variety), size (volume), and deviation in data flow rate

(velocity) are extracted. Based on these data characteristics, the expected workload of
the next time interval is predicted using support vector machine (SVM). Followed by
this, the cloud resource manager dynamically allocates the available cloud resources
depending on the predicted workload. The presentation outcomes have confirmed
that the proposed system has less execution time and achieves better utilization of
resources, when compared to the existing tools.