|
| Study reference | Data model | Resource frameworks | Study features | Evaluation/ranking methodology |
|
| [8] | Data stream processing systems | Storm, Flink, Spark, Samza | A brief comparison of resource frameworks | ✘ |
|
| [9] | Batch and stream processing systems | Horizontal scaling systems, such as peer-to-peer, MapReduce/MPI, and Spark, and vertical scaling systems, such as CUDA and HDL | Comparison of horizontal and vertical scaling systems | Theoretical comparison of resource frameworks |
|
| [10] | Batch and stream processing engines | MapReduce, Spark, Flink, and Storm as well as machine learning libraries | Machine learning libraries and their evaluation mechanism | Performance comparison with respect to machine learning toolkits |
|
| [11] | Batch and stream processing frameworks | Hadoop, Storm, and other big data frameworks | In-depth analysis of big data opportunities and challenges | ✘ |
|
| [12] | Batch and stream processing frameworks | Hadoop, Spark, Storm, Flink, and Tez as well as SQL, Graph, and bulk synchronous parallel model | Analysis of current open research challenges in the field of big data and the promising directions for future research | ✘ |
|
| [13] | Stream processing engines | Apache Storm, S4, Flink, Samza, Spark Streaming, and Twitter Heron | Classification of elasticity metrics for resource allocation strategies that meet the demands of stream processing services | Evaluation of elasticity/scaling metrics for stream processing systems |
|
