Efficient ELM-Based Two Stages Query Processing Optimization for Big Data

Abstract EN

MapReduce and its variants have emerged as viable competitorsfor big data analysis with a commodity cluster of machines.

As anextension of MapReduce, ComMapReduce realizes the lightweight communicationmechanisms to enhance the performance of query processingapplications for big data.

However, different communication strategies ofComMapReduce can substantially affect the executions of query processingapplications.

Although there is already the research work that can identifythe communication strategies of ComMapReduce according to thecharacteristics of the query processing applications, some drawbacks stillexist, such as relative simple model, too much user participation, and relativesimple query processing execution.

Therefore, an efficient ELM-basedtwo stages query processing optimization model is proposed in this paper,named ELM to ELM (E2E) model.

Then, we develop an efficient sampletraining strategy to train our E2E model.

Furthermore, two query processingexecutions based on the E2E model, respectively, Just-in-Time executionand Queue execution, are presented.

Finally, extensive experiments areconducted to verify the effectiveness and efficiency of the E2E model.