Research Area:  Big Data

Abstract:

With the exponential increase of the data amount in the past years, data analytics and data processing became essential to any organization. As Moore-s law has been exceeded since several years ago, the excessive data hides indeed highly useful information. The real challenge is to successfully extract the information using an effective process and with a reasonable cost. Therefore, various processing techniques have emerged. Indeed, big data processing methods can be classified into several types like batch based, stream based, Graph based, DAG based, interactive based and visual based. All data processing techniques follow the same cycle: data collection, data preparation, data input, processing, data output/interpretation and data storage. Although having this similarity, these approaches have certainly different use cases, architectures and tools. This paper focuses on two types, namely: Batch-based processing and stream-based processing. After defining these two approaches, a comparative study is conducted and some key features are highlighted.

Keywords:  
Big Data
Data Source
Data Insights
Batchbased Processing
Stream-based Processing

Author(s) Name:  Sarah Benjelloun; Mohamed El Mehdi El Aissi; Yassine Loukili; Younes Lakhrissi; Safae Elhaj Ben Ali; Hiba Chougrad; Abdessamad El Boushaki

Publisher name:  IEEE

DOI:  10.1109/ICDS50568.2020.9268684

Volume Information:  

Paper Link:   https://ieeexplore.ieee.org/abstract/document/9268684